Quantifying the Adhesion and Interaction Forces Between

HA's are considered the most abundant OM in many types of groundwater (8), and have been used to represent NOM in laboratory studies (9-11). Recent ...
0 downloads 0 Views 334KB Size
Environ. Sci. Technol. 2007, 41, 8031–8037

Quantifying the Adhesion and Interaction Forces Between Pseudomonas aeruginosa and Natural Organic Matter LAILA I. ABU-LAIL,† YATAO LIU,‡ ARZU ATABEK,‡ AND T E R R I A . C A M E S A N O * ,‡ Department of Civil and Environmental Engineering and Department of Chemical Engineering, Life Sciences and Bioengineering Institute at Gateway Park, Worcester Polytechnic Institute, 100 Institute Rd, Worcester, MA, USA 01609

Received May 04, 2007. Revised manuscript received September 10, 2007. Accepted September 11, 2007.

Atomic force microscopy (AFM) was used to characterize interactions between natural organic matter (NOM), and glass or bacteria. Poly(methacrylic acid) (PMA), soil humic Acid (SHA), and Suwannee River humic Acid (SRHA), were adsorbed to silica AFM probes. Adhesion forces (Fadh) for the interaction of organic-probes and glass slides correlated with organic molecular weight (MW), but not with radius of the organic aggregate (R), charge density (Q), or zeta potential (ζ). Two Pseudomonas aeruginosa strains with different lipopolysaccharides (LPS) were chosen: PAO1 (A+B+), whose LPS have common antigen (A-band) + O-antigen (B-band); and mutant AK1401 (A+B-). Fadh between bacteria and organics correlated with organic MW, R, and Q, but not ζ. PAO1 had lower Fadh with silica than NOM, which was attributed to negative charges from the B-band polymers causing electrostatic repulsion. AK1401 adhered stronger to silica than to the organics, perhaps because the absence of the B-band exposed underlying positively charged proteins. DLVO calculations could not explain the differences in the two bacteria or predict qualitative or quantitative trends in interaction forces in these systems. Molecular-level information from AFM studies can bring us closer to understanding the complex nature of bacterial–NOM interactions.

Introduction Pseudomonas aeruginosa is a Gram-negative bacterium with relevance to a number of environmental applications, including bioremediation (1). This microbe is a common environmental isolate (2) and is an opportunistic pathogen that causes infections of the urinary tract, soft tissue, and respiratory system, particularly in patients with weaker immunity (3, 4). The properties of the lipopolysaccharides (LPS) of P. aeruginosa help determine the bacterial behavior in these applications. Smooth LPS contain lipid A, a core oligosaccharide, and a polysaccharide side chain (termed the O-antigen), whereas strains that lack the O-antigen are * Corresponding author phone-508-831-5380; e-mail - terric@ wpi.edu. † Department of Civil and Environmental Engineering. ‡ Department of Chemical Engineering. 10.1021/es071047o CCC: $37.00

Published on Web 10/12/2007

 2007 American Chemical Society

called “rough”. The LPS structure of Pseudomonas sp. mediates bacterial retention/adhesion to surfaces including glass, polystyrene (5), and soil (6, 7). P. aeruginosa’s behavior in the environment depends on interactions with organic macromolecules (OM) present in the subsurface, broadly termed natural organic matter (NOM), which include humic acids (HA’s). HA’s are considered the most abundant OM in many types of groundwater (8), and have been used to represent NOM in laboratory studies (9–11). Recent experimental advances in nuclear magnetic resonance (12) and atomic force microscopy (AFM) (13), as well as molecular modeling (14), are providing better chemical and structural information on NOM. NOM adsorbs to bacterial surfaces over a wide pH range, which is believed to be due to hydrophobic interactions (8, 15, 16). In laboratory column studies, retention of bacteria decreased on HA-coated substrates (clay and iron-oxidecoated quartz) (17, 18), either by sorption onto the bacterial cell walls, thus increasing their negative charge, or by competition between OM and bacteria for sediment sorption sites (19). Organic coatings can mask the properties of underlying mineral surfaces, causing the transport of Comamonas sp. DA001 to be indifferent to mineral composition (20). Although adsorption studies have provided valuable information, these methods cannot directly quantify forces between bacteria and NOM. To date, only one study directly characterized single-cell bacterial-NOM interactions. The attachment of Escherichia coli K-12 D21 to HA-coated silica or glass was measured with differential electrophoresis, and found to increase when the bulk concentration of HA was greater (21). The primary objective of this work was to use AFM to quantify the interactions between two strains of P. aeruginosa and OM-coated probes, focusing on how LPS structure affects bacterial adhesion and interactions with OM., since no study has directly assessed bacteria–NOM adhesion forces using AFM. Likewise, the role of LPS structure on bacterial–NOM interactions has not been addressed. Since some studies have used simple polymer models to represent NOM, a secondary objective was to evaluate whether PMA could represent the interactions of the more complex forms of OM. The evaluation of forces and characterization of colloids at one ionic strength and pH are meant to provide a single snapshot to show the complex interplay of the different interactions occurring in these systems.

Materials and Methods OM-Coated Surfaces. The NOM compounds investigated were poly(methacrylic) acid (PMA; Polymer Source, Inc.; Dorval (Montreal), Canada; molecular weight (MW) ) 6800 g/mol) and two reference standards from the International Humic Substances Society (St. Paul, MN): Suwannee River Humic Acid (SRHA; MW ) 1490 g/mol (22)) and Soil Humic Acid (SHA; MW ) 2500 g/mol (22)). Solutions (100 mg/L, based on the total dry mass) of PMA, SRHA, or SHA were prepared before each experiment. AFM probes were functionalized with OM using a stepwise process. Silicon nitride cantilevers with attached 1.0 µm silica (SiO2) spheres (Novascan Technologies, Inc.; Ames, IA) were cleaned by exposure to UV irradiation for 40 min, and rinsed with 0.01 M NaOH and ultrapure water (Milli-Q water, Millipore Corp.; Billerica, MA). An intermediate coating step of ironoxide was used to help OM attach to the silica (11). The surface was coated with iron oxide by incubating in 10-5 M FeCl3 solution for 15 h, followed by OM solution for >16 h. VOL. 41, NO. 23, 2007 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8031

Clean glass slides were treated with iron oxide and OM, using the same procedure applied to the colloidal probes. Tapping-mode AFM imaging in water was used to determine the surface coverage by OM (Dimension 3100, Nanoscope IIIa; Veeco Metrology, Woodbury, NY; silicon cantilevers (NSC36-C, Mikromasch; Wilsonville, OR)). The “threshold” function within the AFM software was used to count the density of particles in an image, and to estimate the percentage covered by particles. The average height of the aggregates was used to calculate the radius (R) of each type of OM. To measure the zeta potentials (ζ) of the colloids, the same Fe oxide and OM treatment was applied to 1.0 µm silica microspheres (Polysciences, Inc.). The ζ values were measured six times and averaged, using 1 mM NaCl (Zetasizer Nano ZS, Malvern Instruments; Southborough, MA). Preparation and Characterization of Bacteria. P. aeruginosa wild-type smooth strain PAO1 and semirough mutant strain AK1401 (23) were provided by Professor Gerald Pier (Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital/ Harvard Medical School). (LPS schematic shown in Supporting Information Figure S.1). In addition to the core and lipid A regions, the LPS of PAO1 contains two chemically distinct saccharides, termed A-band (common antigen) and B-band polysaccharides (A+B+) (24). AK1401 does not express B-band polysaccharides (A+B-) (23). Single colonies from Tryptic Soy Agar (40.0 g/L; Sigma) were transferred to Tryptic Soy Broth (30.0 g/L; Sigma) and grown at 37 °C to midexponential phase. Bacteria were washed by centrifugation at 3500g for 15 min. Sessile drop contact angle measurements (θw) were taken with water (Ramé-Hart goniometer 100–00; Netcong, NJ) for bacteria on membrane filters (0.45 µm; Millipore, Billerica, MA). Zeta potentials of bacterial suspensions in 1 mM NaCl were measured six times and averaged. Interaction energy profiles as a function of separation (h) were calculated using Derjaguin–Landau–Verwey–Overbeek (DLVO) theory for sphere–sphere interactions, which included nonretarded (h < 5 nm) and retarded (h > 5 nm) Lifshitz–van der Waals interactions (25), and electrostatic interactions (26), consistent with previous bacterial adhesion studies (27). Parameters used for the calculations are given in Supporting Information Table S.1. AFM Measurements. Force measurements were performed with 1.0-µm colloidal silica probes. Spring constants were 0.118 ( 0.049 N/m (n ) 30), measured using a thermal technique (28). Deflection voltage-separation distance data were converted into force-separation profiles (29). For all force measurements, the scan rate was 1.0 Hz, the scan size was 5.0 µm, and the trigger mode was set to “off”. Bacteria were immobilized on clean glass slides using a cross-linking reaction between amine and carboxyl groups (30). Experiments were performed for 10 cells per condition (n ) 5 per bacterium). All force measurements were performed in 1 mM NaCl solution (pH ∼ 6.3). In approach curves, average values of the force at zero distance (Fo), corresponding to the cantilever deflection prior to the point of reaching the constant compliance region, and decay length values corresponding to zero force (LD), were estimated from 50 data sets per sample. From AFM retraction profiles, the adhesion force corresponding to each adhesion peak between the probe and sample (Fadh) was calculated for 50 retraction profiles. The Mann–Whitney rank sum test was used for all statistical analyses. In control experiments, interaction forces between clean glass slides and bare silica probes or probes modified with OM were captured (n ) 50) in 1 mM NaCl. 8032

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 41, NO. 23, 2007

FIGURE 1. Distribution of Fadh values for coated colloidal probes interacting with glass slide.

FIGURE 2. Representative AFM approach and retraction profiles for interaction of PMA probe with P. aeruginosa AK1401 (1 mM NaCl, pH ) 6.3).

Results Cell-Free Experiments with NOM. For all colloids, including the Fe-oxide and OM surfaces, ζ values were negative (Supporting Information Table S.2), which showed that the iron oxide layer did not completely coat the glass. The surface with the most negative potential was uncoated glass (ζ ) -90.27 mV). The corresponding AFM force profiles always showed evidence of repulsion in the approach curves (examples of approach curves are shown in Supporting Information Figure S.2). This was true even for the ironoxide coated probe, since the coverage of iron was presumably too low to mask the negative charge of silica. Although the coverage of the OM on the AFM tips was not complete, the force profiles are highly reproducible, suggesting that the large radius of the sphere allowed for an averaging effect of the colloidal forces measured. After verifying the tip modification was reproducible, we concentrated our analysis on the adhesion forces gathered from the retraction portions of AFM force cycles (Figure 1). PMA showed the weakest Fadh with glass, with >75% of adhesive forces