pubs.acs.org/Langmuir © 2010 American Chemical Society
Salt Modulates Bacterial Hydrophobicity and Charge Properties Influencing Adhesion of Pseudomonas aeruginosa (PA01) in Aqueous Suspensions Jacob J. Shephard,† David M. Savory,† Phil J. Bremer,‡ and A. James McQuillan*,† †
Department of Chemistry and ‡Department of Food Science, University of Otago, P.O. Box 56, Dunedin, New Zealand Received December 20, 2009. Revised Manuscript Received April 7, 2010
The influence on cell hydrophobicity of differential extension with ionic strength of lipopolysaccharide molecules (LPS), which exist as charged and uncharged polymers at the surface of the Gram-negative bacterium Pseudomonas aeruginosa (PA01), has been investigated. Attenuated total reflection infrared (ATR-IR) spectral absorptions from a single layer of cells adsorbed to ZnSe increased in intensity with increasing NaCl concentration up to 0.1 mol L-1. Dynamic contact angle measurements (Wilhelmy plate tensiometry) made with a ZnSe plate having an adsorbed cell layer and the adherence of the cells to hexadecane suggest that PA01 cells were most hydrophobic in contact with 0.1 mol L-1 NaCl solutions. These data indicate a charge screening induced compression of the charged LPS polymers decreasing the cell-surface approach distance and increasing the cell hydrophobicity due to the greater surface predominance of the uncharged LPS polymers. Interestingly, adsorbed cell layers in 0.3 mol L-1 NaCl had a lower IR absorption intensity, and PA01 cells suspended in 0.3 mol L-1 were found to be more hydrophilic, indicating that other factors influence the cell-surface approach distance and hydrophobicity. The examination of cell electrophoretic mobility variation with NaCl concentration suggests that the compression of charged polysaccharides increases the polysaccharide charge density and may also reduce the flow of liquid through the polysaccharide layer affecting the effective potential at the interface, the cell hydrophobicity, and the cell-surface approach distance.
Introduction The adhesion and self-aggregation of many Gram-negative bacterial species has generally been modeled using DLVO theory,1 where pH and ionic strength are the main variables. However, the adhesion kinetics predicted from the estimated electrostatic and dispersive interactions has not been in agreement with experimental results.1 Many reasons have been suggested for this lack of agreement including the incorrect estimation of surface potential due to the permeability of the surface layers to liquid flow, regulation of surface charge in response to electrolyte concentration, and the influence of acid-base and steric interactions.2-9 A property of Gram-negative bacteria which has not yet been considered is the impact of ionic strength on the extension of charged surface lipopolysaccharide (LPS) molecules and how this may influence cell surface charge properties and hydrophobicity. A greater understanding of the influence of ionic strength on these cell surface properties may help explain deviations from DLVO-based adhesion rate predictions and lead to better strategies for control of bacterial attachment to surfaces. *Departments of Chemistry and Food Science, University of Otago, P.O. Box 56, Dunedin, New Zealand. (1) Poortinga, A. T.; Bos, R.; Norde, W.; Busscher, H. J. Surface Sci. Rep. 2002, 47, 1–32. (2) Sonohara, R.; Muramatsu, N.; Ohshima, H.; Kondo, K. Biophys. Chem. 1995, 55, 273–277. (3) Duval, J. F. L.; Busscher, H. J.; Belt-Gritter, B.; van der Mei, H.; Norde, W. Langmuir 2005, 21, 11268–11282. (4) Morisaki, H.; Nagai, S.; Ohshima, H.; Ikemoto, E.; Kogure, K. Microbiology 1999, 145, 2797–2802. (5) Dan, N. Colloids Surf., B 2003, 27, 41–47. (6) Busalmen, J. P.; de Sanchez, S. R. J. Ind. Microbial. Biotechnol. 2001, 26, 303–308. (7) Bayouda, S.; Othmane, A.; Bettaieb, F.; Bakhrouf, A.; Ouda, H. B.; Ponsonnet, L. Mater. Sci. Eng. 2006, 26, 300–305. (8) Rijnaarts, H. H. M.; Norde, W.; Lyklema, J.; Zehnder, A. J. B. Colloids Surf., B 1999, 14, 179–195. (9) Hong, Y.; Brown, D. G. Langmuir 2008, 24, 5003–5009.
Langmuir 2010, 26(11), 8659–8665
The surfaces of Gram-negative species are decorated with LPS, macromolecules with lipid and polysaccharide components. The polysaccharide component extending from the outermost lipid bilayer has a functionality and length dependent on the species and the strain. Pseudomonas aeruginosa, which is an important Gram-negative bacterium due to its role in many human infections, has an ability to respond to environmental influences by modifying its surface polysaccharide functionality to vary the length and relative proportion of two LPS molecules, termed A-band and B-band.10-13 A-band LPS has a polysaccharide component comprising about 20 repeat trisaccharide units of D-rahmnose while B-band LPS has a longer polysaccharide component comprising between 30 and 50 repeat trisaccharide units of aminomanuronic acids and N-acetyl-D-fucosamine.14,15 A-band LPS is electroneutral at physiological pH and fairly hydrophobic due to terminal methyl groups. B-band LPS is negatively charged at physiological pH due to the existence of carboxylate groups and is therefore relatively hydrophilic.13 Cells grown in the planktonic state show an increased proportion of B-band LPS relative to those grown in a biofilm, suggesting that this negatively charged polysaccharide may act to reduce selfaggregation and adhesion to surfaces.16,17 Due to the existence of (10) Rivera, M.; Bryan, L. E.; Hancock, R. E. W.; McGroarty, E. J. J. Bacteriol. 1988, 170, 263–275. (11) Makin, S. A.; Beveridge., A. J. J. Bacteriol. 1996, 178, 3350–3352. (12) Eboigbodin, K. E.; Newton, J. R. A.; Routh, A. F. Appl. Microbiol. Biotechnol. 2006, 73, 669–675. (13) Soares, T. A.; Staatsma, T. P.; Lins, R. D. J. Braz. Chem. Soc. 2008, 19, 312–320. (14) Rocchetta, H. L.; Burrows, L. L.; Lam, J. S. Microbiol. Mol. Biol. Rev. 1999, 63, 523–553. (15) Shephard, J. J.; McQuillan, A. J.; Bremer, P. J. J. Appl. Environ. Microbiol. 2008, 74, 6980–6986. (16) Makin, S. A.; Beveridge, T. J. Microbiology 1996, 142, 299–307. (17) Beveridge, T. J.; Makin, S. A.; Kadurugamuwa, J. L.; Li, Z. FEMS Microbiol. Rev. 1997, 20, 291–303.
Published on Web 04/23/2010
DOI: 10.1021/la1007878
8659
Article
Shephard et al.
these surface polysaccharides, the pH and ionic strength of the suspending solution has a strong influence on the surface properties of Gram-negative cells.18 For example, due to an electrolyte charge-screening effect the extension of the charged surface polysaccharide LPS component into the surrounding solution varies with ionic strength.19,20 The influence of ionic strength and pH on the interaction between Pseudomonas aeruginosa and a ZnSe surface has been probed in an earlier study by this group using attenuated total reflectance infrared (ATR-IR) spectroscopy.21 The absorbance of a typical band relating to adsorbed bacterial cells was found to vary with time as the number of adsorbed cells increased toward complete monolayer coverage and also with the pH and ionic strength of the suspending solution.21 An observed variation in IR band absorbance for an adsorbed cell layer with changes in solution pH and ionic strength was found to be due to variation in the cell-surface approach distance.22 In a later study by Kang et al.,23 the bacterial adsorption process was found to be irreversible, and by fitting absorbance vs time data for a bacterial adsorption experiment, adsorption rate data were obtained independently from information about the average cell-surface approach distance. In the present study, ATR-IR spectroscopy has been used to obtain information about the adhesion rate and approach distance of PA01 cells to ZnSe for ionic strengths between 0.001 and 0.3 mol L-1, a wider range than previously investigated. The focus of the current work is on the relationship between these parameters and measurable surface properties relating to cell surface adhesion such as the cell hydrophobicity. It is hypothesized that the influence of ionic strength on the differential extension of the two polysaccharide LPS components determines the cell-surface approach distance and cell hydrophobicity. The variation in hydrophobicity of PA01 cells with NaCl concentration has been investigated with dynamic contact angle measurements using Wilhelmy plate tensiometry and by the cell adhesion behavior to hexadecane. The surface properties of PA01 have also been investigated by examining electrophoretic mobility data over a range of NaCl concentrations using the soft particle approach.24 The well-correlated results obtained from these techniques highlight the importance of electrolyte composition and ionic strength in determining the surface properties of Gram-negative cells.
Materials and Methods Growth and Harvest of Bacteria. Pseudomonas aeruginosa wild-type strain (PA01) obtained courtesy of Dr. J. S. Lam (University of Guelph, Canada) was maintained as stock cultures in the Microbank cryovial system (Prolab Diagnostics) at -80 °C. The cells were cultured on a Vogel medium comprising 2 g citric acid (Pierce); 3.5 g NaNH4HPO4 3 7H20 (BDH, 99%); 0.2 g MgSO4 3 7H2O (Ajax, analytical reagent); 10 g K2HPO4 (Riedel de Haen, analytical reagent); and 5 g of glucose (May and Baker) per liter of water (Milli-Q, Millipore) used throughout.25 A glucose stock solution was filter sterilized (0.45 μm) and added to the other components previously sterilized by autoclaving. To prepare bacterial suspensions, media (10 mL) were inoculated (18) Hermansson, M. Colloids Surf., B 1999, 14, 105–119. (19) Abu-Lail, N. I.; Camesano, T. A. Langmuir 2002, 18, 4071–4081. (20) Abu-Lail, N. I.; Camesano, T. A. Biomacromolecules 2003, 4, 1000–1012. (21) McWhirter, M. J.; McQuillan, A..J.; Bremer, P. J. Colloids Surf., B 2002, 26, 365–372. (22) McWhirter, M.; Bremer, P. J.; McQuillan, A. J. Langmuir 2002, 18, 1904– 1907. (23) Kang, S. Y.; Bremer, P. J.; Kim, K.; McQuillan, A. J. Langmuir 2006, 22, 286–291. (24) Ohshima, H. Colloids Surf., A 1995, 103, 249–255. (25) Vogel, H. J.; Bonner, D. M. J. Biol. Chem. 1956, 218, 97–106.
8660 DOI: 10.1021/la1007878
with a single colony from a 24 h streak plate and agitated on a shaking table at 170 rpm for 24 h at 25 °C. A 1 mL sample of the resulting bacterial suspension was then added to 250 mL of media and grown for a further 24 h (early stationary phase). The bacterial suspensions were centrifuged (8 min at 8000 g, 4 °C) and the pellet resuspended in 0.03 mol L-1 NaCl, shaken, and washed by centrifugation 2 additional times with water. Bacterial suspensions with an optical density (OD) at 600 nm of 0.5 which corresponds to ∼1 109 mL-1 of colony forming units (CFU) were prepared in solutions containing 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 mol L-1 NaCl at a pH of ∼6. Characterization of ZnSe Surface. For all adsorption processes, the chemical and physical nature of the substrate and the adsorbate are important to consider. There have been few reports on the surface properties of ZnSe in spite of its frequent use in IR spectroscopy addressing surface chemistry. Chemical vapor deposited (CVD) ZnSe is known to slowly oxidize in air to ZnO and SeO2 (SeO32-), which will influence the surface properties of the ATR internal reflection element (IRE) used for bacterial adsorption.26 The salts of the selenite ion (SeO32-) and its oxidation product selenate (SeO42-) are both highly soluble in aqueous solutions, so surface oxidation of the CVD ZnSe (Harrick) will leave a Se-depleted oxide surface. The selenium depletion process is expected to be nonuniform, as oxidation in air at ambient temperatures is accompanied by surface migration and cluster formation.26 The electrophoretic mobility of micrometersized ZnSe particles was measured and found to have a zeta potential between 10 mV at pH 3 and -10 mV at pH 10, indicating the existence of some ionizable functional groups. The presence of some ZnO which has an isoelectric point of ∼8 is a likely explanation for this behavior.27 This conclusion is supported by dynamic water contact angle measurements made using the Wilhelmy plate technique on a clean (as below) rectangular prism of ZnSe from the same source as for the IRE. The ZnSe was found to have an advancing water contact angle of 75° ( 2° and a receding water contact angle of ∼30° ( 2°. While the advancing contact angle was found to be stable with time during a repeat dipping process, the receding contact angle dropped by ∼5° during the first 4 dips (20 min) before becoming constant. These data indicate that the surface was chemically inhomogeneous with both hydrophobic (ZnSe) and hydrophilic (ZnO) regions,28 and also appear to indicate that selenium dioxide (SeO32-) formed from ZnSe oxidation is removed after 20 min in aqueous solution. Thus, the cleaned ZnSe IRE surface has consistent properties after 20 min immersion in aqueous solution and is relatively hydrophobic. Infrared Spectroscopy of PA01 Adhesion to ZnSe. The ZnSe IRE was cleaned immediately before use by light polishing on a microcloth (Buehler) with small amounts of 0.015 μm γ-Al2O3 aqueous slurry and rinsed thoroughly with water. The prism was then kept immersed in water for at least 20 min before measurements were made. For each IR adsorption experiment, bacterial suspensions with NaCl concentrations of 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 mol L-1 prepared from separate cultures were flowed over the surface of a ZnSe IRE with 13 reflections (Horizon, Harrick) and the ATR-IR spectrum recorded (FTS 4000, Digilab) from 64 scans at a resolution of 4 cm-1. After baseline correction of the spectra (Win-IR Pro 3.4, Digilab), correcting to zero absorbance at 1800 cm-1, the absorbance at 1550 cm-1 from the amide II band was used to plot the increase in bacterial absorptions with time and to follow the formation of a single cell layer.23 The variation in absorbance at 1550 cm-1 with time has been found in this work and in prior studies to be reproducible.22,23 For an irreversible adsorption process from a (26) Smathers, J. B.; Keedler, E.; Bennett, B. R.; Jonker, B. T. Appl. Phys. Lett. 1998, 72, 1238–1240. (27) Kosmulski, M. Surface charging and points of zero charge, Surfactant Science Series, Vol. 145; CRC: Boca Raton, 2009, p 214. (28) Adamson, W. A.; Gast, A. P. Physical chemistry of surfaces, 6th ed.; Wiley: New York, 1997; pp 355-362.
Langmuir 2010, 26(11), 8659–8665
Shephard et al. constant concentration flowing suspension, the variation in fractional surface coverage with time is expected to follow a 1- e-kt relationship where k is the pseudo-first-order adsorption rate constant and t is the adsorption time.29 For typical ATR-IR spectroscopy adsorption experiments, absorbance of an adsorbed species is proportional to fractional surface coverage, and can be used to extract kinetic data. For this series of adsorption experiments, it is assumed that at constant ionic strength the average distance of an adsorbed cell from the ZnSe surface is constant with changes in coverage. A nonlinear curve fitting procedure (OriginPro 8.0, OriginLab) was used to fit the absorbance vs time data and extract adsorption rate parameters. Hydrophobicity of Pseudomonas aeruginosa (PA01). Advancing and receding contact angles of a rectangular shaped prism of ZnSe with dimensions of 4 10 40 mm having an adsorbed single layer of PA01 were recorded using the Wilhelmy plate technique and an automated tensiometer (DCA 100, First Ten Angstroms).The ZnSe was cleaned and prepared as for the IR experiments, then immediately immersed for 2 h at a plate speed of 0.1 mm s-1 in a bacterial suspension with an optical density of 0.5 and a NaCl concentration of 0.03 mol L-1. Loosely attached cells were removed by twice dipping the ZnSe plate having adsorbed cells in a 0.03 mol L-1 NaCl solution. To make dynamic water contact angle measurements on the adsorbed cells, the coated ZnSe plate was twice dipped at a plate speed of 0.1 mm s-1 in aqueous solutions with stepwise increasing NaCl concentrations of 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 mol L-1. To calculate advancing and receding contact angles, the tension measurements recorded on the second dip were extrapolated to zero immersion (Camtell software) using literature values (between 72.4 and 72 0.9 mN m-2) for the liquid-vapor interfacial tension of NaCl solutions at 22 °C.30,31 The adherence of PA01 cells to hexadecane droplets was measured with a similar methodology to that reported elsewhere but with the use of larger volumes.32 20 mL of bacterial suspensions with an OD of 0.5, with a NaCl concentration of 0.001, 0.003, 0.01, 0.03, 0.1, and 0.3 mol L-1 and pH adjusted to 4, was combined with 10 mL of hexadecane (Aldrich) in 6 identical 50 mL separating funnels. The separating funnels were shaken using a wrist arm shaker for 1 min and left to separate for 20 min. The aqueous layer was then removed and the OD of eight 1.0 mL samples recorded using a 48 well plate and plate reader (Synergy 2, Biotech). The bacterial suspensions were then removed from the wells and returned to separating funnels before the process was repeated. ODs were recorded after 1, 2, and 3 min of shaking. The reduction in OD after subsequent shaking time was used to calculate the percentage of cells adsorbed to the hexadecane droplets.
Electrophoretic Mobility of Pseudomonas aeruginosa (PA01). The PA01 cell suspension used to measure the electrophoretic mobility was prepared from a single culture which was washed as above, then diluted with water or a NaCl solution to make a suspension with a OD of ∼0.05 (approximately 1/10 of that used for IR experiments) and final NaCl concentrations of ∼0, 0.001, 0.005, 0.01, 0.025, 0.05, 0.1, 0.2, and 0.3 mol L-1. The pH of the resulting bacterial suspensions was ∼6. Electrophoretic mobility measurements (Zetasizer Nano ZS, Malvern) at 25 °C were made within 2 h of sample preparation. The results given are the average from three independently grown cultures. Conductivity measurements confirmed that very little ion flux between cells and solution had occurred in the 2 h prior to measurement of the samples.
Results and Discussion IR Analysis of Adsorption Kinetics of PA01 Bacteria to ZnSe. Bacterial adsorption experiments were carried out with the (29) Young, A. G.; McQuillan, A. J. Langmuir 2009, 25, 3538–3548. (30) Peterson, P. B.; Saykally, R. J. Annu. Rev. Phys. Chem. 2006, 57, 333. (31) Matubayasi, N.; Matsuo, H.; Kazuo, Y.; Shin-ichiro, Y.; Asako, M. J. Colloid Interface Sci. 1999, 209, 398–402. (32) Rosenburg, M. FEMS Microbiol. Lett. 1984, 289–295.
Langmuir 2010, 26(11), 8659–8665
Article
Figure 1. ATR-IR spectra of PA01 adsorbed to ZnSe from a bacterial suspension with an optical density of 0.5 (∼1 109 CFU mL -1) and a NaCl concentration of 0.1 mol L-1 after 1, 2, and 3 h. Background for the spectra is from 0.1 mol L-1 NaCl solution over ZnSe.
same concentration of suspended bacteria (∼1 109 mL -1) but with different NaCl concentrations. In each experiment, an appropriate NaCl solution without bacteria was initially flowed over the ZnSe surface to obtain a background spectrum for the subsequent spectra of adsorbed bacteria resulting from flow of a bacterial suspension.23 Figure 1 shows typical data with IR absorptions in the 1800-900 cm-1 spectral region observed after 1, 2, and 3 h flow of a PA01 suspension in 0.1 mol L-1 NaCl. The wavenumber and band shape of absorptions in this region are similar to those reported for many bacterial strains and arise from the internal components of the bacteria, the cell wall, and the bacterial surface appendages. The most prominent absorptions are the amide I and amide II bands of proteins peaking at ∼1650 and ∼1550 cm-1, respectively. Assignments for other Pseudomonas aeruginosa bands have been previously reported.23 The shape and peak wavenumber of adsorbed PA01 bands did not vary significantly with NaCl concentration or time. For all adsorption experiments, the intensity of bacterial absorptions increased rapidly at the onset of adsorption but less rapidly as time passed. Although this pattern was similar for all NaCl concentrations, the initial rate of absorption increase and maximum absorbance reached varied for experiments carried out at different NaCl concentrations. Absorbance at 1550 cm-1 was chosen to indicate the increasing bacterial coverage with time, because this strong amide II spectral band is not influenced by water-related absorptions which perturb the amide I band. The changes in absorbance at 1550 cm-1 with time for all bacterial adsorption experiments are shown in Figure 2. There is a marked variation in the measured absorbance vs time plots over the NaCl concentrations investigated, as previously noted.22 These data appear to indicate that the adhesion behavior of the bacteria was strongly influenced by the ionic strength of the aqueous suspension. Additionally, the data trends indicate that the maximum IR absorbance expected for each studied NaCl concentration is significantly different. PA01 adsorption to ZnSe follows an irreversible adsorption relationship23 and the line associated with the 0.1 mol L-1 NaCl data points in Figure 2 is an example of such a fit of the absorbance data. The results of the fitting of all of the Figure 2 absorbance data to the irreversible adsorption relationship Amax(1 - ekt) are given in Table 1 containing the pseudofirst-order adsorption rate constants (k) for a ∼1 109 mL-1 bacterial suspension and the predicted absorbance corresponding to saturation coverage (Amax) for each NaCl concentration. The absorbance at saturation coverage (Amax) for PA01 on ZnSe DOI: 10.1021/la1007878
8661
Article
Shephard et al.
Figure 2. Absorbance at 1550 cm-1 in the spectra of PA01 during adsorption to ZnSe from bacterial suspensions with an OD of 0.5 (∼1 109 CFU mL-1) and NaCl concentrations of 0.001 (0), 0.003 (4), 0.01 (O), 0.03 (9), 0.1 (2), and 0.3 mol L-1(b). The line represents the fit of the absorbance data for 0.1 mol L-1 NaCl solution to the irreversible adsorption relationship (Amax(1 - ekt)). Table 1. Adsorption Rate Parameters Derived from Fitting Absorbance vs Time Data to Amax(1 - ekt) for the Adsorption of PA01 to ZnSe at Different Ionic Strengths c/mol L-1
104 k/s-1
Amax
R2
0.001 0.003 0.01 0.03 0.1 0.3
5.0 6.3 2.5 3.5 2.3 1.9
0.009 0.010 0.021 0.034 0.066 0.027
0.981 0.984 0.998 0.997 0.999 0.987
corresponds closely to a complete single layer of cells.23 The best fit of the absorbance data to the irreversible adsorption relationship was observed for a NaCl concentration of 0.1 mol L-1 with an R2 value of 0.999. The quality of the fit was reduced at higher or lower NaCl concentrations but was above an R2 of 0.98 for all bacterial adsorptions. The good fit obtained for bacterial adsorption carried out at NaCl concentrations between 0.001 and 0.3 mol L-1 indicates that the adsorption rate constants provide a satisfactory basis for comparison of the adsorption rate data with Amax corresponding to surface saturation. The decreased quality of fit at high and low NaCl concentrations may indicate that bacterial adsorption from these solutions was less irreversible than for the bacterial adsorption carried out at 0.1 mol L-1. The fitted Amax values and adsorption rate constants (k) are independent variables and will now be considered separately. Amax was found to vary considerably with NaCl concentration as anticipated from the Figure 2 data. Amax increases from an absorbance of ∼0.01 to ∼0.07 for a NaCl concentration increasing from 0.001 to 0.1 mol L-1, then decreases to ∼0.03 for a NaCl concentration of 0.3 mol L-1. Amax is determined by the distance between the bacterial cells and the ZnSe surface (cell-surface approach distance),33 and also by the surface packing density for saturation coverage. Related ionic strength influences have been found in the adsorption behavior and coverage trends of charged colloidal particles on different surfaces.34 A similar variation with ionic strength in absorbance of a Pseudomonas aeruginosa monolayer adsorbed on ZnSe has been previously reported, and (33) Harrick, N. J. Internal reflection spectroscopy; Wiley: New York, 1967; p 43. (34) Yuan, Y.; Oberholzer, M. R.; Lenhoff, A. M. Colloids Surf., A 2000, 165, 125–141.
8662 DOI: 10.1021/la1007878
Figure 3. Schematic diagram illustrating the influence of ionic strength and LPS extension on the cell surface density at saturation coverage and cell-surface distance with both quantities influencing the predicted IR absorbance at saturation coverage (Amax).
effluent analysis indicated that no change in coverage had occurred.22 The variation in absorbance with ionic strength was attributed to a change in the cell-surface approach distance. The similar absorbance data obtained in the present work suggests that Amax is primarily determined by the cell-surface approach distance with the exponential decay of IR radiation with distance from the ZnSe surface providing a major influence.33 The cellsurface approach distance and surface packing density at saturation coverage are influenced by the extension of the negatively charged polysaccharide LPS component which is primarily determined by ionic strength.19,21-23 Figure 3 illustrates this graphically. Cells adsorbed from a low ionic strength solution (A) with extended surface polysaccharides have a lower surface density at saturation coverage and approach the ZnSe IRE less closely than cells with compressed surface polysaccharides adsorbed from a higher ionic strength solution (B). Using the simple model shown in Figure 3 with spherical cells of radius 2 μm, a 7-fold increase in Amax corresponds to about a 30% reduction in the cell radius or a 650 nm reduction in the extension of the surface polymers. The surface polysaccharides present on a similar bacterium, Pseudomonas putida, were found by AFM to have a maximum extension of 850 nm.19 After the observed increase in Amax with increasing ionic strength, the decreased value of Amax found for a NaCl solution concentration of 0.3 mol L-1 cannot easily be explained by a charge screening effect. Interestingly, a similar decrease in absorbance with increasing ionic strengths between 0.03 and 0.15 mol L-1 was observed in an earlier study of PA01 adsorption to ZnSe,22 and a transition between a trend of increasing to decreasing bacterial adhesion of Pseudomonas aeruginosa35 and Vibrio alginolyticus1 cells with ionic strength increasing above 0.1 mol L-1 has been previously reported. This adsorption behavior may be due to the influence of electrostatic bacteria-bacteria and bacteria-surface interactions8,18 or due to a change in the cell hydrophobicity. The adsorption rate constant (k) is determined by the strength and range of interactions between the PA01 cells and the ZnSe (35) Stanly, P. M. Can. J. Microbiol. 1983, 29, 1493–1499.
Langmuir 2010, 26(11), 8659–8665
Shephard et al.
surface. It is well-understood that the distance range of electrostatic interactions between the negatively charged cell and weakly positively charged surface is reduced with increases in ionic strength, but the attractive dispersive forces, which primarily drive the adsorption, do not change significantly.18 On the basis of only the attractive charge interactions, the adsorption rate constant is expected to decrease with increases in NaCl concentration as the electrostatic attraction between the cell and surface diminishes. However, this simplistic view which is applicable to some colloidal systems does not take into consideration the deformable polysaccharide structure of the cell surface which is influenced by ionic strength. The adsorption rate constant varied from ∼6.3 10-4 to ∼1.9 10-4 s-1 for NaCl concentrations of between 0.003 and 0.3 mol L-1, and there is a general trend of decreasing k with increasing NaCl concentration, although this is not entirely consistent. Thus, other factors which vary with ionic strength, such as the deformable nature of the cell surface and the cell hydrophobicity, may also have an influence on the adsorption rate trend. Dynamic Water Contact Angles of PA01 Bacteria Adsorbed to ZnSe. In addition to electrostatic interactions which have been addressed in the previous section, the strong hydration of the ZnSe surface or the cell can increase the barrier to adsorption and decrease the adsorption rate and cell surface approach distance. This is because during an inner-sphere adsorption process the hydration shells surrounding the surface and the cell become disrupted, a process which is less favored for strongly hydrated surfaces. The extent to which bacterial cells are hydrated may be expected to vary with NaCl concentration due the greater influence of relatively hydrophobic A-band LPS in higher ionic strength solutions when B-band LPS is compressed. Wilhelmy plate tensiometry was used to test this hypothesis by measuring dynamic water contact angles for a single layer of PA01 cells adsorbed to ZnSe in contact with NaCl concentrations between 0.001 and 0.3 mol L-1. Dynamic water contact angles are influenced by the chemical nature of the surface,28 with strongly hydrated surfaces (hydrophilic) having relatively low water contact angles while weakly hydrated surfaces (hydrophobic) have higher water contact angles. The advancing and receding contact angles of the clean ZnSe IRE prior to adsorption of a cell layer were found to be ∼75° and ∼25°, respectively, and did not vary significantly with NaCl concentration. After adsorption of a cell layer to the ZnSe IRE, the receding contact angle did vary with NaCl concentration. The variation in advancing and receding water contact angles on ZnSe with an adsorbed cell layer is shown in Figure 4. The advancing contact angle does not vary much with NaCl concentration and is similar to the advancing contact angle of water on the clean ZnSe IRE (75°) which suggests that the surface coverage achieved after 2 h of immersion was not a complete single layer. The receding contact angle was found to vary between ∼10° in dilute NaCl solutions and ∼40° for a NaCl concentration of 0.1 mol L-1 indicating that cells were adsorbed to the ZnSe plate and that the hydration of these cells was strongly influenced by the NaCl concentration. The receding contact angle was greatest for a NaCl concentration of 0.1 mol L-1, indicating that cells under these conditions are in their most hydrophobic state with the more hydrophobic A-band LPS polysaccharide component being predominant at the bacterial surface. The receding contact angle observed for cells immersed in 0.3 mol L-1 NaCl solution was lower than that for cells immersed in 0.1 mol L-1 NaCl. This result appears to indicate that the trend of increasing cell hydrophobicity with increasing ionic strength is reversed at ionic strengths greater than ∼0.1 mol L-1. Electrophoretic mobility Langmuir 2010, 26(11), 8659–8665
Article
Figure 4. Advancing (4) and receding (0) contact angles with increasing NaCl concentration recorded using a ZnSe plate with an adsorbed PA01 cell layer.
Figure 5. Adsorption of Pseudomonas aeruginosa (PA01) cells to hexadecane droplets from NaCl solutions with a pH of 4 after 1, 2, and 3 min of shaking. Error bars represent standard deviation from 5 OD measurements.
measurements presented in a later section further probe the interaction between the charged bacterial surface polysaccharides and water. Adhesion of PA01 Cells to Hexadecane. The microbial attachment to hydrocarbon (MATH) assay is a technique often employed to investigate the hydrophobicity of microbial species. Microbial cells have been found to adsorb to the surface of hexadecane droplets reducing the concentration of cells in the aqueous phase. The concentration of PA01 cells in the aqueous phase, measured by optical density, was used to estimate the proportion of cells adsorbed to the hexadecane droplets after 1, 2, and 3 min contact and hence indicate the hydrophobicities of the cells at different NaCl concentrations. The cell hydrophobicity is generally probed at a pH of 4, which is close to the isoelectric point of hexadecane droplets, to reduce the effect of electrostatic interactions on the rate of bacterial adsorption to the droplets.36 Figure 5 shows the percentage of cells adsorbed to the hexadecane droplets from solutions with six NaCl concentrations between 0.001 and 0.3 mol L-1 after 1, 2, and 3 min of shaking. (36) Busscher, H. J.; van de Belt-Gritter, B.; van der Mei, H. Colloids Surf., B 1995, 5, 111–116.
DOI: 10.1021/la1007878
8663
Article
For all NaCl concentrations, the proportion of cells adsorbed to the hexadecane droplets increased after 1 to 2 min shaking time, then decreased slightly or remained unchanged after the third minute. The proportion removed after 1 and 2 min varied with NaCl concentration with the greatest reduction in OD being from cell suspensions with NaCl concentrations between 0.01 and 0.1 mol L-1, suggesting that these cells were the most hydrophobic. The relatively small difference in the proportion of cells adsorbed to the hexadecane droplets after 2 and 3 min indicates that the available hexadecane surface area was saturated with cells after 2 min. The slight decrease in adsorbed cells after 3 min observed for solutions with ionic strengths greater than 0.03 mol L-1 may be due to a reduction in the available hexadecane surface area due to coalescence of the droplets. As the hydrophobicity of the solution cells is influenced by ionic strength and the cells in contact with hexadecane are not, the initial rate of cell adsorption (% adherence after 1 and 2 min shaking time) shows more significant variation with ionic strength than that after 3 min, which corresponds to saturation of hexadecane with cells. These results are in general agreement with those obtained from the dynamic contact angle measurements and suggest that more hydrophobic cells are able to approach the ZnSe surface more closely. It is also significant that the most hydrophobic cells have the best fit to irreversible adsorption behavior (Table 1) indicating that the adsorption of PA01 to ZnSe is strongly influenced by cell hydrophobicity. Electrophoretic Mobility Variation of PA01 with NaCl Concentration. The adhesion of Gram-negative bacteria to a surface may also be influenced by variation in the electrostatic potential at the cell-solution boundary with variation in LPS extension which alters the charge density in the polysaccharide layer. This potential also influences the cell hydrophobicity. The electrophoretic mobility of a solid particle with charges at its surface can be related to its surface potential by the Smoluchowski equation. For such a particle, the electrophoretic mobility decreases with increasing ionic strength tending toward zero at high ionic strengths due to charge shielding by counterions in the electrical double layer. For a Gram-negative bacterium which has negative charge distributed in three dimensions in a polysaccharide layer external to the cell membrane, the electrophoretic mobility does not always decrease to zero with increasing ionic strength.37 Ohshima and co-workers were the first to model the variation in mobility with ionic strength for a particle with a charge distribution of this kind, termed a soft particle.24 They have suggested that for a soft particle such as a Gram-negative bacterium, the density of charges (Ffix) and amount of fluid flow though the polysaccharide layer, termed softness (1/λ), can be found from a plot of electrophoretic mobility against ionic strength. It is assumed in this model that the charge density and softness do not change with ionic strength. The variation with NaCl concentration (equivalent to ionic strength) in electrophoretic mobility for PA01 is shown in Figure 6A. For cells suspended in water, the electrophoretic mobility was found to be ∼ -3 10-8 m2 V-1 s-1 and decreased in magnitude to ∼ -1 10-8 m2 V-1 s-1 as the NaCl concentration increased. The electrophoretic mobility data for PA01 do not tend to zero with increasing NaCl concentration, characteristic of soft particles, due to fluid flow through the polysaccharide layer external to the cell membrane. Also shown in Figure 6A are the electrophoretic mobility values expected for soft particles with charge density of 0.2 C L-1 and a softness of 6 nm (solid line) (37) Tsuneda, S.; Jaekook, J.; Hayashi, H.; Aikawa, H.; Hirata, A; Sasaki, H. Colloids Surf., B 2003, 29, 181–188.
8664 DOI: 10.1021/la1007878
Shephard et al.
Figure 6. (A) Variation in electrophoretic mobility with NaCl concentration for Pseudomonas aeruginosa PA01 (0) at pH ∼6 and lines which represent two possible fits for the PA01 mobility data from soft particle theory. In (B) and (C), either the softness (1/λ) or charge density (Ffix) has been fixed and the other parameter has been varied with NaCl concentration.
and a charge density of 2.5 C L-1 and a softness of 1.5 nm (dashed line). These lines represent two possible fits of the PA01 electrophoretic mobility data. While both lines fit the PA01 mobility data fairly well at ionic strengths of >0.1 mol L-1, each fails to Langmuir 2010, 26(11), 8659–8665
Shephard et al.
fit the data completely at ionic strengths of 1.5 nm the charge density variation goes through a maximum, decreasing at NaCl concentrations of >0.05 mol L-1, indicating an extension of the charged external polysaccharides. The softness also goes through a maximum for charge densities between 0.1 and 2.5 C L-1 but at lower NaCl concentrations. At low NaCl concentrations and for charge densities of >0.5 C L-1, no softness value fits the data. These data suggest that a large increase in charge density with increasing NaCl concentration or a variation in both parameters is required to satisfactorily fit the PA01 mobility data. Ionic strength may be expected to impact the charge density and fluid flow within the external polysaccharides due to the charged polysaccharide (38) Duval, J. F. L.; Busscher, H. J.; Belt-Gritter, B.; van der Mei, H.; Norde, W. Langmuir 2005, 21, 11268–11282. (39) Buhler, E.; Boue, F. Macromolecules 2004, 37, 1600–1610. (40) Ballauff, M. Prog. Polym. Sci. 2007, 32, 1135–1151. (41) Tsuneda, S.; Aikawa, H.; Hayashi, H.; Hirata, A. J. Colloid Interface Sci. 2004, 279, 410–417.
Langmuir 2010, 26(11), 8659–8665
Article
compression/extension. The data appear to show that the parameters vary over different ionic strength ranges with softness more greatly influenced at high ionic strengths, possibly due to the complete retraction of the B-band LPS polysaccharide component to within the A-band polysaccharides at the surface of the cell.
Conclusions ATR-IR spectroscopy was employed to investigate the influence of ionic strength on the rate of PA01 adsorption to ZnSe and the average distance between an adsorbed cell and the ZnSe surface indicated by the maximum absorbance parameter Amax. It has been previously shown that deformable negatively charged polysaccharides, which coat the surface of Gram-negative species, such as Pseudomonas aeruginosa, are less extended with increasing NaCl concentration.19,20 In the present work, it was found that the average distance between a cell and the ZnSe surface decreased with increasing NaCl concentration up to ∼0.1 mol L-1 then increased for a NaCl concentration of 0.3 mol L-1. This result is mainly due to the effect of electrolyte screening on the extension of the charged B-band LPS polysaccharide component. However, the increased distance between the cell and the surface at high ionic strength cannot be explained by this effect alone. In this work, it has been shown that ionic strength also affects cell hydrophobicity with a similar ionic strength trend to that of Amax observed in IR studies of cell adsorption to ZnSe. The dynamic contact angles of a single layer of cells adsorbed to a ZnSe plate and rate at which cells adsorbed to hexadecane droplets suggested that cells in contact with a 0.1 mol L-1 NaCl solution were the most hydrophobic. The increased hydrophobicity of cells in this concentration range has been attributed to the increased predominance of relatively hydrophobic A-band LPS at the surface with B-band LPS polysaccharides more compressed. However, the increased hydrophilicity of PA01 cells in aqueous NaCl solutions with a concentrations of 0.3 mol L-1 cannot be directly due to an electrolyte charge screening effect on the extension of B-band LPS. The variable hydrophobicity of the cells was found to influence the rate and reversibility of their adhesion to ZnSe. The poor fit of PA01 electrophoretic mobility data to the soft particle theory shows that the variable extension of charged polysaccharides may also affect charge density in and fluid flow through the PA01 outermost polysaccharides. A reduced fluid flow through the polysaccharides at ionic strengths of >0.05 mol L-1 due to retraction of B-band LPS to within the A-band polysaccharides may have influenced the cell hydrophobicity. More detailed information about the variation with ionic strength in the charge density of Gramnegative bacteria due to variable extension of charged polysaccharides may allow this hypothesis to be tested. Acknowledgment. This work was supported by the New Zealand Foundation for Research Science and Technology, contract number CO8X0409.
DOI: 10.1021/la1007878
8665
