ARTICLE pubs.acs.org/Biomac
Passive Control of Quorum Sensing: Prevention of Pseudomonas aeruginosa Biofilm Formation by Imprinted Polymers Elena V. Piletska,*,† Georgios Stavroulakis,† Lee D. Larcombe,† Michael J. Whitcombe,† Anant Sharma,‡ Sandy Primrose,† Gary K. Robinson,§ and Sergey A. Piletsky† †
Cranfield Health, Cranfield University, Cranfield, Bedfordshire MK43 0AL, United Kingdom Bedford Hospital, NHS Trust, Kempston Road, Bedford, Bedfordshire MK42 9DJ, United Kingdom § School of Biosciences, University of Kent, Canterbury CT2 7NJ, United Kingdom ‡
ABSTRACT: Here we present the first molecular imprinted polymer (MIP) that is able to attenuate the biofilm formation of the opportunistic human pathogen Pseudomonas aeruginosa through specific sequestration of its signal molecule N-(3-oxododecanoyl)-L-homoserine lactone (3-oxo-C12-AHL). The MIP was rationally designed using computational modeling, and its capacity and specificity and that of a corresponding blank polymer toward signal molecule of P. aeruginosa (3-oxo-C12-AHL) and its analogue were tested. The biofilm formation in the presence of polymers and without polymers was studied using scanning confocal laser microscopy. Staining with crystal violet dye was used for the quantification of the biofilm formation. A significant reduction of the biofilm growth was observed in the presence of MIP (>80%), which was superior to that of the resin prepared without template, which showed a reduction of 40% in comparison with biofilm, which was grown without polymer addition. It was shown that 3-oxo-C12-AHL-specific MIP prevented the development of quorum-sensing-controlled phenotypes (in this case, biofilm formation) from being up-regulated. The developed MIP could be considered as a new tool for the elimination of life-threatening infections in a multitude of practical applications; it could, for example, be grafted on the surface of medical devices such as catheters and lenses, be a component of paints, or be used as a wound adsorbent.
’ INTRODUCTION Bacterial biofilms are a persistent and widespread problem in many sectors including fuel and oil storage, water storage, and food processing equipment. The entrenchment of microorganisms in a mucilaginous exopolysaccharide (EPS) matrix makes them resistant to attack both by chemical antimicrobials (disinfectants and antibiotics) and to elements of the immune system.1,2 Given the latter, it is not surprising that by one estimate, provided by the Centers for Disease Control and Prevention (CDC), 65% of human bacterial infections involve biofilms.1 The difficulty in eliminating microorganisms protected within biofilms involves lengthy treatment cycles, allowing time for drug resistance to develop, compounding the problem. If an efficient strategy aimed at disabling or reversing biofilm formation could be developed (ideally one that does not put evolutionary pressure on the cells to develop resistance), then it would enable one to steer the course of infection away from endangering the host. If the bacteria are therefore prevented from depositing a biofilm and remain in the planktonic state, then the infection will be more easily treated through drug therapies and through control by the immune system. The use of nonadherent coatings on medical devices, such as catheters, is one strategy that can be used; however, the effect is confined to the device and offers no protection to adjacent tissues and surfaces.3 A much better strategy would be to disrupt the regulatory system responsible for initiating biofilm formation, thereby preventing biofilm formation in the first place. r 2011 American Chemical Society
The strategy, which will be exploited in this study, involves the control of the bacterial quorum sensing (QS). QS is a mechanism by which bacteria assess their population density through the secretion and sensing of small signal molecules, the concentration of which in the local environment indicate the number of neighboring cells of the same species. When this concentration exceeds a certain threshold, bacteria are triggered to express phenotypic traits associated with infection, including biofilm formation and the production of virulence factors. Pseudomonas aeruginosa is an example of opportunistic human pathogen that is implicated in the infection of many tissues through biofilm formation, including the lungs of patients suffering from cystic fibrosis, where it is regarded as one of the major causes of mortality associated with the condition.4 P. aeruginosa infections are generally associated with a high degree of mortality, regardless of appropriate antimicrobial treatment regimes.5 In common with other Gram negative bacteria, the signal molecules of P. aeruginosa principally implicated in this pathway are acylated homoserine lactones (AHLs).6 This pathway can be targeted through treatment with compounds that inhibit the production or binding of AHLs to block QS;7-9 however, these proposed drug candidates would need lengthy clinical trials and toxicity testing before they could be approved for clinical use. The Received: November 24, 2010 Revised: February 4, 2011 Published: March 01, 2011 1067
dx.doi.org/10.1021/bm101410q | Biomacromolecules 2011, 12, 1067–1071
Biomacromolecules
Figure 1. Proposed mechanism for the attenuation of biofilm formation by sequestration of quorum signal molecules by a molecular imprinted polymer.
solution that we propose is to use synthetic polymers capable of attenuation of bacterial quorum sensing in P. aeruginosa by sequestration of the signal molecules (Figure 1). This approach has already been explored by our team using a model system based on the marine microorganism Vibrio fischeri.10 It was shown that it was possible to attenuate the QSdependent phenotypes of V. fischeri (i.e., bioluminescence and biofilm formation) by removal of the signal molecule N-(βketocapryloyl)-L-homoserine lactone (3-oxo-C6-AHL) using “rationally-designed” polymers. The presence of polymer within the growth medium had no effect on bacterial growth; removing the possibility of selection pressures leading to mutagenesis and the development of resistance. Here we describe the next step in the development of this technology, which is aimed at designing signal molecule-sequestering polymers for the control of P. aeruginosa infections and to demonstrate their potential for practical medical applications. The compound 3-oxo-C12-AHL was selected as the target for molecularly imprinted polymer (MIP) preparation (through imprinting of the commercially available racemic compound) because it is crucially involved in regulating the expression of extracellular virulence factors11,12 and biofilm formation11,13,14 in P. aeruginosa and has been found to be an important factor in the development of P. aeruginosa infections in animal models.15,16
’ MATERIAL AND METHODS Materials. The wild strain of P. aeruginosa PAO1 was used throughout the study. N-(3-oxododecanoyl)-DL-homoserine lactone (3-oxoC12-AHL) was purchased from GLSynthesis. N-butyryl-L-homoserine lactone (C4-AHL), itaconic acid (IA), ethylene glycol dimethacrylate (EGDMA), 1,10 -azobis(cyclohexane-carbonitrile), N,N-dimethylformamide (DMF), phosphate-buffered saline (PBS, pH 7.4), acetonitrile, and formic acid were purchased from Sigma (Sigma-Aldrich, Gillingham, U. K.). Methanol, ethanol, and crystal violet were obtained from Acros Organics (Fisher, U.K.). Wheat germ agglutinin-Alexa Fluor 488 conjugate (WGA) was purchased from Invitrogen (Carlsbad, CA). Luria-Bertani (LB) broth and agar were from Merck (Poole, Dorset, U.K.). HPLCquality water that was purified using a Milli-Q system was used in all experiments. Computational Modeling. Molecular modeling was undertaken using a workstation from Research Machines running the CentOS 5 GNU/Linux operating system, configured with a 3.2 GHz core 2 duo processor, 4 GB memory, and running the SYBYL 7.3 software suite (Tripos, St. Louis, MO). The LEAPFROG algorithm was applied to screen the library of functional monomers for their possible interactions with the template, resulting in a table ranking the monomers according
ARTICLE
to their binding scores (in kilocalories per mole). The library contained 20 functional monomers that are commonly used in molecular imprinting and possessed polymerizable residues and residues able to interact with a template through ionic and hydrogen bonds, van der Waals, and dipole-dipole interactions.17 The charges for each atom were calculated using Powell method in combination with Gasteiger-Huckel charges and Tripos force field. The same method was applied to refine the structures of the monomers using energy minimization to a value of 0.01 kcal mol-1. A refining step for optimization of the polymer composition was performed.18 It involved a molecular dynamics simulation of the prearrangement of the itaconic acid monomer around the template prior to polymerization. This was carried out by saturating the space around the template with monomers in a defined box that was heated to 600 K and slowly cooled to 300 K. Energy minimization was carried out to 0.05 kcal mol-1. At the end of the simulation, the number and the position of the functional monomers were examined. The type and quantity of the monomers participating in the complex with the template (first shell layer), or monomers that interact with the first shell layer of monomers, stabilizing complex (second shell layer) indicate the type and ratio of the template and monomers in an optimized MIP composition. Polymer Preparation. MIP composition consisted of 5 mg of 3-oxo-C12-AHL (16 μmol) as a template, 25 mg of functional monomer (IA) (190 μmol), 0.47 g of cross-linker (EGDMA) (2.37 mmol), 0.5 g of DMF, and 5 mg of 1,1-azobis (cyclohexanecarbonitrile). The composition of the corresponding control (blank polymer) was prepared similar to MIP but without template. The polymer mixtures were purged with nitrogen and polymerized for 12 h at þ80 °C. After synthesis, all polymers were ground and sieved to obtain the fractions with size 38125 μm. Polymers were thoroughly washed using Soxhlet extraction with methanol to remove unreacted monomers and dried. To ensure that no template remained in the MIP, the polymer was thoroughly washed in Soxhlet using methanol, followed by regular changes of acidic methanol, acidic water, water, and methanol. Washing continued for 3 days until the concentration of the template in the eluent measured using HPLC-MS became below quantification level (>2 ng mL-1).
Quantification of 3-oxo-C12-AHL and its Analogue Using HPLC-MS. To quantify the sequestration of 3-oxo-C12-AHL and its analogue C4-AHL, the HPLC-MS method was optimized. HPLC separation was conducted using a Waters 2975 HPLC system equipped with Luna C18 (2) column (150 3 mm, 3 μm, Phenomenex). Mobile phase: (A) water and (B) methanol in a binary system with 0.1% of formic acid as an additive. The elution gradient: linear gradient from 50% methanol/0.1% formic acid to 100% B from 0 to 10 min, then for 5 min solution B (100% methanol/0.1% formic acid), and return to 50% methanol at 20 min (total time 20 min). The flow rate was 0.2 μL min-1, the injection volume was 10 μL, and the column temperature was þ20 °C. The 102 m/z fragment that is common to both tested AHLs was detected in positive ion mode using a mass spectrophotometer (Micromass Quatro Micro and ESI interface, Waters).19 The MS parameters were the following: desolvation gas 850 L h-1, cone gas 50 L h-1, capillary 4.5 kV, cone 25 V, CE 20, source temperature þ120 °C, desolvation temperature þ350 °C, collision energy 25 V, and multiplier 650 V. The quantification of 3-oxo-C12-AHL was conducted using MassLynx 4.0 SP software (Waters). Equilibrium Binding Study. An evaluation of the binding capacity of the polymers toward 3-oxo-C12-AHL was carried out as described by L€ubke et al. with minor modifications.20 Samples (10 mg) of either MIP or the corresponding blank polymer were suspended in 1 mL of 3-oxo-C12-AHL solutions (0.1, 1, 5, 10, 25, 50, 75, and 100 μg mL-1 in 20% acetonitrile). The vials were incubated overnight on an orbital shaker (Yellowline OS 2 basic shaker, Fisher) at 200 rpm at 20 °C. After incubation, polymer suspensions were filtered through syringe filters with pore size of 0.20 μm. The concentration of free AHL was assessed by HPLC-MS, and the amount of bound 3-oxo-C12-AHL was calculated by 1068
dx.doi.org/10.1021/bm101410q |Biomacromolecules 2011, 12, 1067–1071
Biomacromolecules
ARTICLE
Figure 2. Structure of the complex between 3-oxo-C12-AHL and five molecules of IA, as predicted by molecular dynamics (SYBYL Vn. 7.3, Tripos). (a) Predicted complex, hydrogen bonds are marked as yellow dotted lines, and a molecule of IA that forms the second shell is marked as *. (b) Structures of 3-oxo-C12-AHL, C4-AHL, and IA. subtraction. The binding characteristics were obtained by curve-fitting the binding isotherms for each section of the isotherm plot using a single-site binding model (Grafit package, Erithacus Software). The level of binding and cross-reactivity of MIP and the corresponding control polymer for C4-AHL were evaluated using the equilibrium binding method, as described above.
Static Biofilm Assay and Laser Scanning Confocal Microscopy (LSCM). The biofilm study was carried out as described by De Kievit et al. with some minor modifications.14 A single colony of P. aeruginosa PAO1 was inoculated in sterile LB broth and incubated overnight at 37 °C with reciprocal shaking at 200 rpm. To prepare the microscope slides with biofilms, we put 1 mL of an overnight culture diluted 1:20 in LB in 4 mL of capacity screw-top vials containing glass slides (25 mm 7 mm) and, where required, 5, 10, and 20 mg of sterile MIP or corresponding blank polymer. The vials were closed and incubated at 37 °C for 24 h under static conditions. Glass slides that contained biofilms were rinsed with sterile water and placed in vials containing 1 mL of 5 μg mL-1 WGA in PBS. After incubation for 2 h at 4 °C in the dark, the slides were removed from the solution and rinsed with sterile water. Biofilms, which were formed on the glass slides in the presence of MIP and blank polymers and without polymers, were visualized using a Zeiss LSM 510 confocal microscope (Zeiss, Germany) equipped with an argon laser at excitation wavelength of 488 nm (green fluorescence) and long pass emission filter of 505 nm. Images were captured using a Plan-Neofluar 40/1.3 oil DIC objective, pinhole set to 1 Airy unit (66 μm), with z-stacks of 1 μm thickness, and processed using the Zeiss LSM software. Quantification of Biofilms. The quantification of biofilm growth was done using staining with crystal violet. The assay was performed as described by Schaber et al.21 Glass slides were carefully removed from each vial (static biofilm formation), rinsed with sterile water, and placed in new vials containing 1 mL of 1% crystal violet solution in water. After incubation at 20 °C for 30 min, the slides were removed from the crystal violet solution, washed with sterile water, dried, and placed in new vials containing 1 mL of 95% ethanol. To quantify how much biofilm was formed, the crystal-violet-stained glass slides were incubated at 20 °C for 1 h, and solution was measured using spectrophotometer UVPC 2100 (Shimazdu, Japan) at wavelength 600 nm.
’ RESULTS AND DISCUSSION Computational Screening. The MIP composition was selected by in silico screening of a library of 20 functional
monomers for their ability to form energetically favorable complexes with an energy-minimized model of the target. The leading candidate monomers were acrylamide, methacrylic acid, itaconic acid (IA), and N,N-methylene-bis-acrylamide, with relative binding energies of -40.78, -32.08, -31.73, and -31.39 kcal mol-1 respectively. Of these four, IA showed a three-point interaction with the target, which would be advantageous for MIP formation. Molecular dynamics predicted that 3-oxo-C12AHL could form a complex with up to 5 mol equiv of IA, four in direct contact with the template and one in the secondary shell (Figure 2); therefore, a template:monomer ratio of 1:5 was used in the preparation of a rac-3-oxo-C12-AHL MIP. A nonimprinted (blank) control polymer was also prepared in the same manner but without the addition of template. MIP and blank characteristics were determined by nitrogen porisimetry as following: MIP specific surface area, 374.2 m2 g-1; total pore volume, 0.425 cc g-1; and average pore diameter, 28.1 Å. The corresponding figures for blank were 303.9 m2 g-1, 0.526 cc g-1, and 41.4 Å. Equilibrium Binding Studies. The binding characteristics of the MIP and blank polymers toward template were determined by batch binding experiments, varying the concentration of AHL from 0.1 to 100 μg/mL (correspondingly, 330 nM to 330 μM). The concentration of 3-oxo-C12-AHL typically found in cultures of P. aeruginosa lies in the range of 200 nM to 10 μM6, although levels as high as 600 μM may be found in the vicinity of a biofilm.22 MIP was found to possess a higher affinity (Kd = 145 ( 2 μM) than blank (Kd = 312 ( 3 μM) as well as a higher binding capacity (22.5 μmol g-1 for MIP as opposed to 14.1 μmol g-1 for blank (the standard deviations for the measurements were below 5%)). Selective binding for the template was demonstrated by incubation with N-butyryl-homoserine lactone (C4-AHL), which is a structural analogue of the template and is the second signal molecule secreted by P. aeruginosa. Binding of C4-AHL by the MIP was below detectable limits, whereas binding by blank was also reduced (0.3 μmol g-1). These results confirmed that the MIP possessed both sufficient specificity and binding capacity to be suitable for a practical demonstration of inhibition of biofilm formation. Static Biofilm Assay. MIP and blank were evaluated for their ability to suppress biofilm formation in a static biofilm assay,14 considered to be a simple and reliable method in the study of 1069
dx.doi.org/10.1021/bm101410q |Biomacromolecules 2011, 12, 1067–1071
Biomacromolecules
ARTICLE
biofilms produced in the presence of the polymers were characterized by more sparse surface distribution of cells and noticeable reduction in the formation of an EPS matrix compared with biofilm formed in the absence of polymer (Figure 3a,d). It is necessary to highlight that the bacterial growth, which was monitored routinely, was not inhibited by the presence of the polymers either by killing bacteria or by depletion of nutrients. To confirm the qualitative observations, above, the extent of biofilm formation was assessed by staining with crystal violet.21 Using this quantitative test, which stains all components of the biofilm, including bacterial cells, showed that addition of 20 mg mL-1 MIP resulted in a reduction in all components of the biofilm by ∼80%, relative to the control; whereas only 40% reduction was seen in the presence of blank polymer (Figure 3g).
’ CONCLUSIONS The described results show that polymeric adsorbents and MIPs imprinted with an appropriate target quorum signal molecule, in particular, are promising candidates for clinical use in the prevention of biofilm formation. The ready incorporation of these polymers, for example, through surface grafting to medical devices (catheters, dialysis membranes, contact lenses, etc.) would improve their resistance to biofilm formation and the resultant medical complications related to P. aeruginosa infections. Whereas it has been demonstrated here for a specific organism, it should be emphasized that this approach is a generic one and can be adapted to a wide range of microorganisms and applications. An added advantage is that MIPs will be effective at suppressing biofilms on all surfaces in the immediate environment, not just on the treated surface, as seen with nonadhesive coatings.3 ’ AUTHOR INFORMATION Corresponding Author Figure 3. Effect of polymers on biofilm formation by P. aeruginosa grown for 24 h on glass slides in the presence of 20 mg mL-1 of polymer. Surface distribution of biofilm: (a) in control; (b) in the presence of blank polymer; and (c) in the presence of MIP; scale bars: 50 μm. Corresponding biofilm thickness shown by z-stack: (d) in control; (e) in the presence of blank polymer; and (f) in the presence of MIP; scale bars (d-f) are: 20, 10, and 10 μm, respectively. (g) Quantitative determination with crystal violet, and error bars represent the standard deviation for six replicates.
QS-controlled biofilm formation. Biofilms were visualized following staining with fluorescent wheat germ agglutinin-Alexa Fluor 488 conjugate (WGA) using laser scanning confocal microscopy (LSCM). WGA is known to bind specifically to Nacetylglucosaminyl and sialic acid residues, both components of the EPS matrix of P. aeruginosa biofilms.23-25 The addition of 20 mg mL-1 MIP, suspended in the culture medium, was shown to completely suppress biofilm formation (Figure 3c,f). Blank polymer also suppressed biofilm development, although to lesser degree (Figure 3b,e). The testing, which was made with smaller quantities of the polymers (5 and 10 mg), also demonstrated the reduction of the biofilm formation but to a lesser extent than in the presence of 20 mg of the polymer. The percentage of the inhibition of biofilm in the presence of 5 mg of MIP was 70% (55% for 5 mg of blank polymer) and in the presence of 10 mg of MIP was 75% (65% for 10 mg of blank polymer) when compared with biofilm formation in the absence of the polymer. The
*Tel: þ44(0)1234 758325. E-mail: e.piletska@cranfield.ac.uk.
’ ACKNOWLEDGMENT G.S. would like to thank the Bedford Hospital and British Council in Germany for financial support of his studentship. ’ REFERENCES (1) Costerton, J. W.; Stewart, P. S.; Greenberg, E. P. Science 1999, 284, 1318–1322. (2) Govan, J. R.; Deretic, V. Microbiol. Rev. 1996, 60, 539–574. (3) Caro, A.; Humblot, V.; Methivier, C.; Minier, M.; Salmain, M.; Pradier, C.-M. J. Phys. Chem. B 2009, 113, 2101–2109. (4) Moreau-Marquis, S.; Stanton, B. A.; O’Toole, G. A. Pulm. Pharmacol.Ther. 2008, 21, 595–599. (5) Rossolini, G. M.; Mantengoli, E. Clin. Microbiol. Infect. 2005, 11, 17–32. (6) Pearson, J. P.; Gray, K. M.; Passador, L.; Tucker, K. D.; Eberhard, A.; Iglewski, B. H.; Greenberg, E. P. Proc. Natl. Acad. Sci. U.S.A. 1994, 91, 197–201. (7) Geske, G. D.; Wezeman, R. J.; Siegel, A. P.; Blackwell, H. E. J. Am. Chem. Soc. 2005, 127, 12762–12763. (8) Hoffmann, N.; Lee, B.; Hentzer, M.; Rasmussen, T. B.; Song, Z.; Johansen, H. K.; Givskov, M.; Hoiby, N. Antimicrob. Agents Chemother. 2007, 51, 3677–3687. (9) Hentzer, M.; Riedel, K.; Rasmussen, T. B.; Heydorn, A.; Andersen, J. Bo.; Parsek, M. R.; Rice, S. A.; Eberl, L.; Molin, S.; Hoiby, N.; Kjelleberg, S.; Givskov, M. Microbiology 2002, 148, 87–102. 1070
dx.doi.org/10.1021/bm101410q |Biomacromolecules 2011, 12, 1067–1071
Biomacromolecules
ARTICLE
(10) Piletska, E. V.; Stavroulakis, G.; Karim, K.; Whitcombe, M. J.; Chianella, I.; Sharma, A.; Eboigbodin, K. E.; Robinson, G. K.; Piletsky, S. A. Biomacromolecules 2010, 11, 975–980. (11) Smith, R. S.; Iglewski, B. H. Curr. Opin. Microbiol. 2003, 6, 56–60. (12) Muh, U.; Hare, B. J.; Duerkop, B. A.; Schuster, M.; Hanzelka, B. L.; Heim, R.; Olson, E. R.; Greenberg, E. P. Proc. Natl. Acad. Sci. U.S.A. 2006, 103, 16948–16952. (13) Davies, D. G.; Parsek, M. R.; Pearson, J. P. Science 1998, 280, 295–298. (14) De Kievit, T. R.; Gillis, R.; Marx, S. Appl. Environ. Microb. 2001, 67, 1865–1973. (15) Teiber, J. F.; Horke, S.; Haines, D. C.; Chowdhary, P. K.; Xiao, J.; Kramer, G. L.; Haley, R. W.; Draganov, D. I. Infect. Immun. 2008, 76, 2512–2519. (16) Tang, H. B.; DiMango, E.; Bryan, R.; Gambello, M.; Iglewski, B. H.; Goldberg, J. B.; Prince, A. Infect. Immun. 1996, 64, 37–43. (17) Piletsky, S. A.; Karim, K.; Piletska, E. V.; Day, C. J.; Freebairn, K. W.; Legge, C.; Turner, A. P. F. Analyst 2001, 126, 1826–1830. (18) Piletska, E. V.; Romero-Guerra, M.; Chianella, I.; Karim, K.; Turner, A. P. F.; Piletsky, S. A. Anal. Chim. Acta 2005, 542, 111–117. (19) Morin, D.; Grasland, B.; Vellee-Rehel, K.; Dufau, C.; Haras, D. J. Chromatogr., A. 2003, 1002, 79–92. (20) L€ubke, C.; L€ubke, M.; Whitcombe, M. J.; Vulfson, E. N. Macromolecules 2000, 33, 5098–5105. (21) Schaber, J. A.; Carty, N. L.; McDonald, N. A.; Graham, E. D.; Cheluvappa, R.; Griswold, J. A.; Hamood, A. N. J. Med. Microbiol. 2004, 53, 841–853. (22) Charlton, T. S.; de Nys, R.; Netting, A.; Kumar, N.; Hentzer, M.; Givskov, M.; Kjelleberg, S. Environ. Microbiol. 2000, 2, 530–541. (23) Goldstein, I. J.; Hayes, C. E. Adv. Carbohydr. Chem. Biochem. 1978, 35, 127–340. (24) Wozniak, D. J.; Wyckoff, T. J. O.; Starkey, M.; Keyser, R.; Azadi, P.; O’Toole, G. A.; Parsek, M. R. Proc. Natl. Acad. Sci. U.S.A. 2003, 100, 7907–7912. (25) Strathmann, M.; Wingender, J.; Flemming, H. C. J. Microbiol. Methods 2002, 50, 237–248.
1071
dx.doi.org/10.1021/bm101410q |Biomacromolecules 2011, 12, 1067–1071
