pubs.acs.org/Langmuir © 2010 American Chemical Society
Surface Nanocrystallization for Bacterial Control Bin Yu,† Adam Lesiuk,‡ Elisabeth Davis,§ Randall T. Irvin,§ and D. Y. Li*,†,‡ †
‡
Department of Biomedical Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2V2, Department of Chemical and Materials Engineering, University of Alberta, Edmonton, Alberta, Canada T6G 2V4, and §Department of Medical Microbiology & Immunology, University of Alberta, Edmonton, Alberta, Canada T6G 2H7 Received March 2, 2010. Revised Manuscript Received April 16, 2010
Stainless steel is commonly used in indwelling medical devices, food preparation, and heavy industry. Bacteria display reduced adherence to nanocrystallized stainless steel. In this article, we present quantitative information on the surface adhesive force, surface electron work function, and bacterial adherence to surfaces of nanocrystallized stainless steel with differing grain sizes. Surface nanocrystallization was achieved by sandblasting followed by recovery treatment. The adhesive force of bacterial binding to nanocrystallized surfaces was measured using an atomic force microscope with a synthetic-peptide-coated AFM tip designed to mimic the bacterial binding site of Pseudomonas aeruginosa, a common pathogen known to form biofilms. The electron work function of the steel surfaces was measured, and bacterial binding assays were performed using subinoculated P. aeruginosa cultures. It was demonstrated that for nanograined steel surfaces, the adhesive force, peptide adherence, surface electron activity, and bacterial binding all decreased with decreasing grain size.
1. Introduction Nosocomial infections are always a concern when indwelling medical devices are utilized. The infections are caused in part by the colonization of the surface of these implants by pathogenic bacteria forming biofilms. The bacteria composing these organized microbial communities are protected such that they become recalcitrant to both the human immune system and antibiotics and have been known to play an important role in the spread of antibiotic resistance.1,2 Because postinfection treatment is difficult and in severe cases treatment requires surgical methods and replacement of the implant,3 cost-effective preventative alternatives are desired. In addition to its effects in medicine, biofilms have been known to cause problems in many other fields. Biofilms that build up on hulls of ships increase drag, thereby escalating fuel consumption. Microbial growth on the heat exchangers of power stations can cause significantly reduced power output. In the food industry, the buildup of biofilms on food-preparation surfaces can cause food contamination despite regular cleaning with disinfectants.3,4 A common pathogen found in these biofilms is Pseudomonas aeruginosa.5 The nature of this bacterium and its adherence to stainless steel, a common metallic material for biomedical and other biorelated applications, has long been a research subject *To whom all correspondence should be addressed. E-mail: dongyang.li@ ualberta.ca. (1) Allison, D. G.; Mcbain, A. J.; Gilbert, P. In Community Structure and Co-operation in Biofilms; Allison, D. G., Gilbert, P., Lappin-Scott, H. M., Wilson, M., Eds.; Cambridge University Press: Cambridge, U.K., 2000; pp 309-327. (2) Costerton, J. W.; Montanaro, L.; Arciola, C. R. Int. J. Artif. Organs 2005, 11, 1062–1068. (3) Carpentier, B.; Cerf, O. J. Appl. Bacteriol. 1993, 75, 499–511. (4) Hood, S. K.; Zottola, E. A. Int. J. Food Microbiol. 1997, 37, 145–153. (5) Costerton, J. W.; Stewart, P. S.; Greenberg, E. P. Science 2009, 284, 1318–1322. (6) Yu, B.; Giltner, C. L.; van Schaik, E. J.; Bautista, D. L.; Hodges, R. S.; Audette, G. F.; Li, D. Y.; Irvin, R. T. J. Bionanosci. 2007, 1, 73–83. (7) Yu, B.; Davis, E. M.; Hodges, R. S.; Irvin, R. T.; Li, D. Y. Nanotechnology 2008, 19, 335101. (8) Stanley, P. M. Can. J. Microbiol. 1983, 29, 1493–1499.
10930 DOI: 10.1021/la100859m
attracting considerable attention.6-10 The type IV pilus (T4P), a filamentous protein structure found at the poles of this bacterium, is responsible for biofilm formation and adherence to both biotic and abiotic surfaces such as stainless steel.11 It is also known that the adhesion of this pilus along with bacterial adhesion in general is significantly affected by the surface free energy of a material.12,13 This free energy, which is in large part due to the surface electronic behavior, may be characterized by the electron work function (EWF), which is the minimum energy (eV) required to move an electron from inside a material to just outside its surface in a vacuum.14,15 Thus, because the EWF is related to the free energy of a surface, it should be a strong predictor of the biofilm-forming propensity of the surface. An effective way to modify the surface free energy of a material without significantly changing its overall chemical composition and bulk properties is through surface nanocrystallization. By performing severe plastic deformation treatments such as ultrasonic shot peening16 or sandblasting treatments,14,17 high-density dislocation cells can be produced in the surface layer of a material, and with low-temperature recovery treatments, these dislocation cells can rearrange into nanosized grains with sharp grain boundaries.18 Although it would be expected that electron activity may be increased as the grain boundary density increases in nanograined steel, because it is known that the electron activity and (9) Vanhaecke, E.; Remon, J.-p.; Moors, M.; Raes, F.; de Rudder, D.; van Peteghem, A. Appl. Environ. Microbiol. 1990, 788–795. (10) Blenkinsopp, S. A.; Khoury, A. E.; Costerton, J. W. Appl. Environ. Microbiol. 1992, 58, 3770–3773. (11) Audette, G. F.; Hazes, B. J. Nanosci. Nanotechnol. 2007, 7, 2222–2229. (12) Busscher, H. J.; Weerkamp, A. H.; Van der Mei, H. C.; Van Pelt, A. W. J.; De Jong, H. P.; Arends, J. Appl. Environ. Microbiol. 1984, 48, 980–983. (13) Fang, H. H. P.; Chan, K. Y.; Xu, L. C. J. Microbiol. Methods 2000, 40, 89–97. (14) Guan, X. S.; Dong, Z. F.; Li, D. Y. Nanotechnology 2005, 16, 2963–2971. (15) Ashcroft, N. W.; Mermin, N. D. Solid State Physics; Saunders: Philadelphia; 1976. (16) Liu, G.; Lu, J.; Lu, K. Mater. Sci. Eng. 2000, A286, 91–95. (17) Wang, X. Y.; Li, D. Y. Wear 2003, 255, 836–845. (18) Mao, X. Y.; Li, D. Y.; Fang, F.; Tan, R. S.; Jiang, J. Q. Philos. Mag. Lett. 2010, 90, 349–360.
Published on Web 04/30/2010
Langmuir 2010, 26(13), 10930–10934
Yu et al.
Article
Figure 1. Sandblasted surfaces annealed at different temperatures for different average grain sizes: (a) 31 ( 1.8, (b) 42 ( 4.5, (c) 50 ( 5.5, (d) 85 ( 6.4, and (e) 105 ( 8.7 nm. (f) The original or microcrystalline stainless steel (annealed without sandblasting) had its average grain size equal to 35 ( 6 μm. Note that the scanned square areas in a-d, e, and f, are 1, 2, and 400 μm in length, respectively.
therefore bacterial adhesion are elevated at grain boundaries,6,19 this is not the case as already documented in passive steels.19 Specifically, in nanograined stainless steel, the high-density grain boundaries promote chromium diffusion, which when reacting readily with surface oxygen forms a protective film (Cr2O3), making the steel surface more stable and resistant to corrosion.17,19 Although this phenomenon has already been preliminarily illustrated and it has been shown that P. aeruginosa has a reduced binding affinity for nanograined steel as compared to its microcrystalline counterpart,7 it has not been documented as to how these properties change as the grain size changes on the nanometer scale. The objective of this work was to evaluate quantitatively how the surface activity and bacterial binding change with the grain size in nanograined stainless steel. These properties were investigated through quantitative analysis by EWF, the standard surface adhesive force (the adhesion measured by a regular silicon nitride AFM tip), and peptide adhesion testing using a previously de novo designed heterodimeric coiled-coil system with an embedded PAK(128-144)ox peptide, which has been shown to mimic the native T4P binding site in the bacterial pathogen P. aeruginosa in order to approximate the adhesive force between the bacterial cells and the stainless steel surface.7 Bacterial binding assays were also performed on microcrystalline and nanograined surfaces to determine bacterial colonization trends on the steel surface with viable bacteria.
2. Materials and Methods Circular samples of stainless steel 304 with 13 mm diameter and 5 mm height were used; they were annealed at 800 °C for 45 min in an Ar atmosphere to remove residual strain. The samples were (19) Li, D. Y. Mater. Soc. Symp. 2006, 0887-Q05-03.1.
Langmuir 2010, 26(13), 10930–10934
polished with sand papers of increasing grit up to no. 1200 and then sandblasted under a pressure of 230 kPa at a 90° impact angle for a period of 20 min. Five sandblasted samples were then annealed at 200, 225, 250, 300, and 400 °C, respectively, for a period of 90 min in an Ar atmosphere to allow dislocations to move and thus turn dislocation cells into nanosized grains. The samples were then lightly polished with sand papers of no. 1200 grit to flatten their surfaces and remove any accumulated thermal oxide on the surface. For comparison, we also prepared an as-received microcrystalline sample that experienced annealing at 300 °C for 90 min without prior sandblasting treatment (denoted as “original” with its average grain size determined to be 35 ( 6 μm, Figure 1f) and a sandblasted sample without annealing, which were polished with sand papers of no. 1200 grit. After light polishing, the samples were then etched for about 20 s in a 50% (v/v) HCl/HNO3 solution with H2O (1:1:1). The surfaces were then imaged with their grain sizes determined using an atomic force microscope (Digital Instruments). The adhesive force of the sample surfaces was measured using AFM with a standard silicon nitride tip on a cantilever having its spring constant equal to 0.06 N m-1. The measurements were taken in at least 10 different locations throughout each sample, and average values were recorded. To measure the peptide adhesive force, the silicon nitride tip was sputter coated with an Au layer of 20 nm thickness. The tip was then immersed in a 25 μM K-coil peptide with an additional N-terminal cystein residue in PBS pH 7.2 for 45 min. The tip was then washed with distilled H2O for 5 min, immersed in 5 μM cysteine in PBS pH 7.2 for 45 min, washed again with distilled H2O for 5 min, immersed in 1 μg/mL of E-coil-PAK(128-144)ox for 45 min, and finally dried in air. The result was a heterodimeric coiled-coil tip with the PAK(128-144)ox receptor binding domain located at the end of the AFM cantilever. This coating method was previously tested to display binding properties similar to those of the native T4P binding site of P. aeruginosa, and the purity of the tip was DOI: 10.1021/la100859m
10931
Article
Yu et al.
Table 1. Different Synthetic Peptides and Their Corresponding Sequences Used in the Preparation of the Derived AFM Peptide Tip peptide
sequence
PAO(128-144)ox E-coil cys-K-coil E-coil PAK(128-144)ox
Ac-A-C-K-S-T-Q-D-P-M-F-TP-K-G-C-D-N-OH Ac-(E-V-S-A-L-E-K)5-OH Ac-C-(K-V-S-A-L-K-E)5-OH H5-(E-V-S-A-L-E-K)5-K-C-T-SD-Q-D-E-Q-F-I-P-K-G-C-S-K
confirmed using HPLC reverse-phase chromatography and mass spectroscopy.20 The peptide sequences used in this experiment are listed in Table 1. The EWF of the surfaces was measured using a scanning Kelvin probe (KP technologies) with an Au tip of 2 mm diameter. Samples were tested in 16 different areas, each of which consisted of 10 10 points for measurement. The values reported are averages of all data points for a sample in comparison with that of Au as the reference. All values including deviations are reported in electron volts (eV). For bacterial binding assays, sample surfaces were washed using 50 mL of 95% (v/v) EtOH for 15 min with gentle agitation and rinsed with distilled water. Immediately before use, all samples were washed for 1 min in 20 mL of acetone with gentle agitation and rinsed with distilled water. The samples were placed into individual wells of a sterile polystyrene Costar six-well cell culture plate (Corning Incorporated). Overnight cultures of P. aeruginosa PAK pilT mutant were subinoculated into 20 mL of prewarmed Luria-Bertani broth (LB) containing 50 μg/mL of tetracycline. Cultures were incubated until an OD600 of ∼0.2 was reached. The cultures were centrifuged at 10 000g for 3 min. Bacterial pellets were washed twice with sterile phosphatebuffered saline (PBS) (10 mM sodium phosphate buffer at pH 7.4 containing 150 mM NaCl) and were resuspended to a final OD600 of 0.6. Seven milliliters of PBS was placed into each well to cover each sample, and 0.5 mL of resuspended bacteria was added to the PBS. Samples were incubated for 1 h at room temperature with gentle agitation. The samples were washed six times with 10 mL of distilled water and were stained in 10 mL of 1 M acridine orange stain for 3 min. Samples were briefly rinsed with distilled water and were observed using a Leitz Laborlux K microscope equipped with a 40 Neoflour lens and epifluorescent illumination. Fifty images per sample were captured using a Canon EOS Rebel XS digital camera, and the number of bound P. aeruginosa per 40 field of view was enumerated.
3. Results and Discussion Figure 1 illustrates AFM images of differently treated samples captured for grain-size determination. The grain size of each sample is an average of 10 measurements from different areas observed under the atomic force microscope. Average values of grain size and standard deviation for different specimens are presented in Table 2. The average grain sizes of the samples annealed in Ar at different temperatures increased from 31 to 105 nm as the annealing or recovery temperature was raised from 200 to 400 °C. For comparison, the average grain size of the asreceived steel after annealing at 300 °C was also measured, which was 35 ( 6 μm. We also examined a sandblasted sample without annealing, which showed its average grain size or, more accurately, itsnanocellular size to be around 63 nm. It needs to be indicated that without annealing (i.e., recovery treatment) a nanocellular structure resulting from severe plastic deformation does not represent nanosized grains but dislocation cells with diffuse boundaries between adjacent cells.18 (20) Giltner, C. L.; Van Schaik, E. J.; Audette, G. F.; Kao, D.; Hodges, R. S.; Hassett, D. J.; Irvin, R. T. Mol. Microbiol. 2006, 59, 1083–1096.
10932 DOI: 10.1021/la100859m
Table 2. Grain Sizes of Samples after Sandblasting and Annealing in an Ar Atmosphere at Different Temperatures for 90 Minutes annealing treatment 200 °C 225 °C 250 °C 300 °C 400 °C unannealed as-received steel, annealed at 300 °C
grain size 31 ( 1.8 nm 42 ( 4.5 nm 50 ( 5.5 nm 85 ( 6.4 nm 105 ( 8.7 nm 63 ( 6 nm 35 ( 6 μm
Adhesive forces of different sample surfaces were measured using a PAK(128-144)ox-peptide-coated AFM tip and are illustrated in Figure 2. As shown in Figure 2a, the adhesive force decreased with the grain size and the attained experimental values are 23.0 ( 2.2, 21.6 ( 1.9, 18.0 ( 1.6, 17.4 ( 1.7, and 16.0 ( 1.7 nN for samples having average grain sizes of 105, 85, 50, 42, and 31 nm, respectively. This decrease in adhesion implies that there is a decrease in the strength of bacterial biofilm binding as the grain size decreases in nanograin stainless steel because the peptide-coated tip has the same amino acid sequence as the PilA binding site on the T4P of P. aeruginosa cells. The original or microcrystalline stainless steel sample had an adhesive force of 26.1 ( 2.4 nN. Thus nanocrystallization had a significant effect on peptide adhesion, where the smallest grain size had the lowest adhesive force as measured using the coiled-coil peptide tip. The sandblasted-without-annealing sample (denoted as S.B.) showed the highest overall adhesive force of 27.0 ( 2.5 nN. This is attributed to the fact that sandblasting generates only a nanocellular structure or a dislocation network that differs from a nanocrystalline structure as the former contains interior dislocations and diffuse boundaries between cells and the latter has nearly defect-free nanocrystals with sharp grain boundaries. The results obtained in the bacterial binding assays were consistent with the aforementioned results, as shown in Figure 2b. The nanocrystallized samples had markedly smaller numbers of bound bacterial cells, compared to those of unannealed (S.B.) and microcrystalline (original) stainless steel samples. A general trend in increased bacterial binding with increased grain size is shown for the nanograined samples. The scattering data of the S.B. sample resulted from its higher surface inhomogeneity caused by sandblasting. Adhesive force testing was also performed using the standard silicon nitride tip; results for nanocrystallized samples are reported in Figure 3. The average adhesive force was 16.0 ( 1.6, 16.2 ( 1.5, 13.0 ( 1.2, 10.1 ( 1.1, and 10.0 ( 1.0 nN for samples having average sizes of 105, 85, 50, 42, and 31 nm, respectively. The adhesive forces of the original stainless steel sample (as received and annealed) and the sandblasted sample (no annealing) were significantly higher than those of the sandblasted and annealed samples, with values equal to 18.8 ( 1.2 and 20.2 ( 1.6 nN, respectively. The adhesive or attractive force of the steel surfaces for the standard tip is lower than that for the peptidecoated tip, but general trends in the variation in adhesive force with grain size are similar. Electron work functions of the samples were measured to correlate the observed adhesion behavior and bacterial binding to the activity of surface electrons, which is related to surface inertness to bacterial binding and biofilm formation. A higher electron work function implies a greater voltage (or applied energy) required to bring electrons from inside a metal at the Fermi level to just outside the metal, corresponding to a more electrochemically stable state. Results of the EWF measurement Langmuir 2010, 26(13), 10930–10934
Yu et al.
Article
Figure 2. (a) Adhesive force of the peptide-coated coiled-coil tip with the PAK(128-144)ox residue designed to mimic the native T4P binding of P. aeruginosa cells. Here the peptide adhesion decreases with grain size for the annealed samples, which indirectly illustrates a trend toward decreased bacterial binding with grain size. (b) Bacterial binding of P. aeruginosa cells to the surface of the stainless steel samples, as enumerated under a 40 objective lens with epifluorescent illumination. The bars overhead illustrate ostatistical differences in data of p < 0.01. Grain sizes corresponding to different annealing temperatures are given in Table 2. S.B. denotes the sandblated sample without annealing, Original denotes the as-received sample experienced annealing.
Figure 3. Adhesive force of the stainless steel samples using a regular AFM tip. The adhesive force (nN) was obtained by multiple testing, with error bars representing the standard deviation of the test data.
are presented in Figure 4. The general increase in the EWF as the grain size decreases for the annealed samples indicates an increase in surface stability. Also, similar to previous results, the microcrystalline stainless steel sample and the unannealed S.B. sample had smaller EWFs, which indicates that the electrons in the surfaces were more readily accessible or more likely to interact with the surrounding medium. The apparent decreases in adhesion, peptide adhesion, and bacterial binding with a decrease in grain size and an increase in the EWF for nanograin steel are consistent and expected. Although the electron activity is known to increase at grain boundaries, in nanocrystalline stainless steel overall chromium diffusion is considerably promoted because of its high-density grain boundaries along which atomic diffusion is accelerated. When chromium reacts with oxygen, it forms a protective layer or passive film on a surface that acts as a relatively inert barrier to the surrounding environment. As the grain size decreases, corresponding to an increase in the grain boundary density, the promoted atomic diffusion helps to reduce the formation of vacancies or pores, forming a more compact passive film with a Langmuir 2010, 26(13), 10930–10934
Figure 4. EWF of the stainless steel samples, measured in electron volts. Increased electron activity is characterized by a lower EWF value. The unannealed sandblasted sample has the lowest EWF, and the annealed 200 °C sample has the highest EWF.
higher Cr concentration,19 which would more effectively block the interaction between surface electrons and the surrounding medium or increase the surface inertness or stability. This study provides quantitative information regarding how specific grain boundary density or changes in grain size on the nanometer level could affect the stability of the surface. The sandblasted but unannealed sample, which superficially appeared to be nanograined under AFM, did not correspond to this described trend because it had a significant difference in adhesion, bacterial binding, and EWF as compared to its annealed counterparts. This sample had a much higher adhesive force measured using both the standard and peptide-coated AFM tip, the number of P. aeruginosa bacteria per field of view was much greater on average, and the EWF was significantly lower, all of which are implications of the potential for increased biofilm formation. As mentioned earlier, without annealing, a sandblasted surface had a nanocellular structure consisting of dislocation cells and dislocations also existed inside the cells, which DOI: 10.1021/la100859m
10933
Article
increased the electron activity and rendered surface more active, as demonstrated by its lower EWF shown in Figure 4. The key factor responsible for the reduced adhesion, bacterial binding, and higher EWF of nanocrystallized surfaces was the formation of a more protective passive film on nanocrystalline surfaces, which formed faster and likely contained more Cr19 that enhanced the passivation capability. Nanocrystallization also influences the interfacial bonding between the substrate and passive film, thus affecting its resistance to mechanical and electrochemical attack.15,16 The passive film on a nanocrystalline substrate may develop inward-growing oxide pegs at high-density grain boundaries, resulting in stronger interfacial bonding between the substrate surface and exterior passive film compared to that of microcrystalline and sandblasted surfaces, where the grain boundary density is lower or effective inward-growing pegs are unlikely to form. It was also noticed that the grain size varied with the annealing temperature in a roughly linear fashion. This demonstrates that with the chosen method of sample preparation stainless steel surfaces could be modified using this cost-effective process with low-temperature annealing or recovery in order to increase the surface stability for reduced biofilm formation. This inexpensive surface-modification technique could effectively be used to sup-
10934 DOI: 10.1021/la100859m
Yu et al.
press the formation of bacterial biofilms not only in medicine but also in food production and others sectors involving biocorrosion and bacterial contamination.
4. Conclusions The electron activity, characterized by the EWF, was lowered as the grain size decreased in nanocrystallized sample surfaces, corresponding to a decrease in the surface adhesive force. Because the derived coiled-coil PAK(128-144)ox peptide contains the same amino acid sequence as and is known to duplicate the binding affinity of the bacterial binding site in P. aeruginosa, a common bacterium compromising biofilms, the decrease in the peptide adhesive force suggests a decrease in the binding force of bacteria to the stainless steel surface. Consequently, the increased inertness of nanocrystallized surfaces led to a decrease in attracting bacteria as observed in the bacterial binding assays. The increase in the surface inertness of nanocrystallized stainless steel surfaces to bacterial binding is attributed to its highdensity grain boundaries, which accelerate atomic diffusion and modify the Cr content in the passive film,19 leading to a decrease in the activity of surface electrons interacting with the surrounding medium.
Langmuir 2010, 26(13), 10930–10934
