pubs.acs.org/Langmuir © 2009 American Chemical Society
Nanomechanical Properties of Dead or Alive Single-Patterned Bacteria Aline Cerf,* Jean-Christophe Cau, Christophe Vieu, and Etienne Dague* CNRS, LAAS, 7 avenue du colonel Roche, F-31077 Toulouse, France, and universit e de Toulouse, UPS, INSA, INP, ISAE, LAAS, F-31077 Toulouse, France Received December 4, 2008 The main goal of this paper is to probe mechanical properties of living and dead bacteria via atomic force microscopy (AFM) indentation experimentations. Nevertheless, the prerequisite for bioAFM study is the adhesion of the biological sample on a surface. Although AFM has now been used in microbiology for 20 years, the immobilization of microorganisms is still challenging. Immobilizing a single cell, without the need for chemical fixation has therefore constituted our second purpose. Highly ordered arrays of single living bacteria were generated over the millimeter scale by selective adsorption of bacteria onto micrometric chemical patterns. The chemically engineered template surfaces were prepared with a microcontact printing process, and different functionalizations of the patterns by incubation were investigated. Thanks to this original immobilization strategy, the Young moduli of the same cell were measured using force spectroscopy before and after heating (45 °C, 20 min). The cells with a damaged membrane (after heating) present a Young modulus twice as high as that of healthy bacteria.
Introduction A major concern in bacteriology is to determine whether a bacterium is dead or alive. Especially when cells are starved, or have been subjected to an oxidative stress1 (e.g., in tap water networks), it is very difficult to make the difference between a cell that is unable to grow and a dead cell. Several techniques were developed to overcome this problem, such as methods based on the ability of a fluorophore to penetrate the bacterial cell wall. Nevertheless, the question of viability is still under debate, and we present, in this paper a new way to probe Escherichia coli viability, using an atomic force microscope (AFM) conducted in the force spectroscopy mode. Thanks to the AFM, operated in liquid, it is now possible to explore living cells at the nansoscale. Nanoindentation experiments have been applied on human platelets,2 living macrophages,3 and red blood cells4 as well as on diatoms5 and different bacteria or fungi.6-11 In 2007, Cross et al.12 showed that mechanical analysis can distinguish cancerous cells from normal ones, even when the cells show similar shape. They found a Young modulus of 1.97 ( 0.70 kPa on benign mesothelial cells, but of only 0.53 ( 0.10 kPa for tumor cells. As far as E. coli is concerned, Eaton et al.13 have recently tested the impact of chitosans (an antibacterial compound) on the elasticity of the cell wall. As the preparation of the AFM sample included a drying step, they *To whom correspondence and requests for materials should be addressed. E-mail:
[email protected] (A.C.);
[email protected] (E.D.). (1) Saby, S.; Leroy, P.; Block, J. C. Appl. Environ. Microbiol. 1999, 65, 5600. (2) Radmacher, M.; Fritz, M.; Kacher, C. M.; Cleveland, J. P.; Hansma, P. K. Biophys. J. 1996, 70, 556. (3) Rotsch, C.; Braet, F.; Wisse, E.; Radmacher, M. Cell. Biol. Int. 1997, 21, 685. (4) Bremmell, K. E.; Evans, A.; Prestidge, C. A. Colloids Surf., B: Biointerfaces 2006, 50, 43. (5) Almqvist, N.; Delamo, Y.; Smith, B. L.; Thomson, N. H.; Bartholdson, A.; Lal, R.; Brzezinski, M.; Hansma, P. K. J. Microsc. 2001, 202, 518. (6) Gaboriaud, F.; Bailet, S.; Dague, E.; Jorand, F. J. Bacteriol. 2005, 187, 3864. (7) Vadillo-Rodriguez, V.; Beveridge, T. J.; Dutcher, J. R. J. Bacteriol. 2008, 190, 4225. (8) Stoica, O.; Tuanyok, A.; Yao, X.; Jericho, M. H. Langmuir 2003, 19, 10916. (9) Schaer-Zammaretti, P.; Ubbink, J. Ultramicroscopy 2003, 97, 199. (10) Touhami, A.; Nysten, B.; Dufr^ene, Y. F. Langmuir 2003, 19, 4539. (11) Dague, E.; Gilbert, Y.; Verbelen, C.; Andre, G.; Alsteens, D.; Dufr^ene, Y. F. Yeast 2007, 24, 229. (12) Cross, S. E.; Jin, Y.-S.; Rao, J.; Gimzewski, J. K. Nat. Nanotechnol. 2007, 2, 780. (13) Eaton, P.; Fernandes, J. C.; Pereira, E.; Pintado, M. E.; Xavier Malcata, F. Ultramicroscopy 2008, 108, 1128.
Langmuir 2009, 25(10), 5731–5736
found Young moduli in the range of 180-220 MPa. In our study, the cells are never allowed to dry, and the force spectroscopy experiments are made in phosphate-buffered saline (PBS) buffer to avoid any electrostatic interactions that can occur in low ionic strength buffers.14 Nevertheless, their nanoindentation experiments revealed mechanical changes in the bacterial cell wall induced by the treatment. In another paper, Volle et al.15 worked on the elasticity of E. coli as it is predated by Bdellovibrio bacteriovorus. They show that the invaded cells are softer (spring constant: 0.064 N/m) than the healthy cells (0.23 N/m). To measure the spring constant, one has to focus on the end of the indentation curve, on high loading forces. The behavior of the cell at high loading forces is known to be caused by osmotic pressure.16,17 As a consequence the modification observed by Volle et al. is essentially due to a dramatic decrease of the cell turgor pressure during the predation process, and not only to the cell wall modification. However, those three studies clearly demonstrate that mechanical properties of cells, which can be probed by force spectroscopy (cell wall elasticity or cell turgor pressure), are specific for a condition (eg.: benign-cancerous, antibacterial or not, predated or not). On this basis we have decided to explore the modifications that could occur in the nanomechanical properties of a single E. coli bacterium, while it is alive and while it is dead. To reach this goal, it has been of first importance to immobilize the living bacteria in an aqueous environment to avoid any cell wall modifications due to a drying step. In the literature, this problem has been overcome by creating positively charged surfaces where negatively charged micro-organisms can be electrostatically immobilized.18 Another solution, very convenient for round shape cells, is to use an Isopore polycarbonate membrane.19 The (14) Gaboriaud, F.; Gee, M. L.; Strugnell, R.; Duval, J. Langmuir 2008, 24, 10988. (15) Volle, C. B.; Ferguson, M. A.; Aidala, K. E.; Spain, E. M.; Nun~ez, M. E. Langmuir 2008, 24, 8102. :: (16) Arnoldi, M.; Bauerlein, E.; Radmacher, M.; Sackmann, E.; Boulbitch, A. Phys. Rev. E. 2000, 62, 1034. (17) Yao, X.; Walter, J.; Burke, S.; Stewart, S.; Jericho, M. H.; Pink, D.; Hunter, R.; Beveridge, T. J. Colloids Surf., B: Biointerfaces 2002, 23, 213. (18) Karreman, R. J.; Dague, E.; Gaboriaud, F.; Quiles, F.; Duval, J. F. L.; Lindsey, G. G. Biochim. Biophys. Acta (BBA): Proteins Proteomics 2007, 1774, 131. (19) Dague, E.; Delcorte, A.; Latge, J. P.; Dufrene, Y. F. Langmuir 2008, 24, 2955.
Published on Web 3/31/2009
DOI: 10.1021/la9004642
5731
Article
Cerf et al.
cells can be mechanically trapped in a pore of the filter and extensively imaged by AFM. Thanks to those methods, a large variety of biophysical experiments (elasticity,20 hydrophobicity,21-23 surface roughness,24 single molecule force spectroscopy25,26) were performed on living cells. Nevertheless, those two methods, as well as mechanical trapping in agarose gel,27 are random. The place where the cell is immobilized is not controlled, and it just sticks randomly on the surface or in the filter pores. For our particular purpose, the ability to position a single bacterium along predefined patterns is a prerequisite. In a previous work,28 we exploited the biotin-streptavidin linkage between biotinlabeled bacteria and streptavidin-coated patterns. Here, we chose to investigate different functionalizations of the patterns to obtain single patterned bacteria but without the need of a chemical fixation. The process can be used either on an oxidized silicon surface for microsystem integration or directly on a glass slide for biological investigations of individual bacteria. In summary, the present paper’s main goal is to probe mechanical properties of living and dead bacteria via AFM indentation experiments. To achieve this goal we overcome a “second” challenge, which is to immobilize a single bacteria cell onto a surface for AFM, without the need of a chemical fixation.
Methods and Materials Surface Chemistry. The principle of the methodology we performed is depicted in Scheme 1. It is based on a conventional microcontact printing (μCP) process and a simple incubation technique to generate functionalized patterns so as to induce local bacteria deposition. The polydimethylsiloxane (PDMS) stamps used for octadecyltrichlorosilane (OTS; 1% v/v) μCP were fabricated according to a previously reported procedure.28 Before printing the OTS molecules, SiO2 substrates were activated with oxygen plasma for 5 min (800 W). The SiO2 patterns defined after OTS μCP process were functionalized with molecules grafted using a simple incubation technique, to be the attractive parts of the surface for bacteria. The deposition of (3aminopropyl)triethoxysilane (APTES; Sigma Aldrich 440140) molecules was carried out by immersion of the substrate into a freshly prepared solution containing APTES molecules diluted at 1% with ethanol for 15 min; the slide was then rinsed with ethanol, dried under nitrogen stream, and heated on a 140 °C plate. Forty-nanometer colloids were incubated on OTSAPTES surfaces for 1 h and then rinsed generously with deionized water. The deposition of epoxide molecules was performed by immersion of the substrate into a freshly prepared solution containing (3-glycidyloxypropyl)trimethoxysilane epoxide molecules (Sigma Aldrich 440167) diluted at 2.5% with ethanol for 30 min, then rinsed with ethanol and dried under a stream of nitrogen. Streptavidin molecules (Sigma Aldrich S3762) were fluorescently labeled with fluorescein thiocarbamoyl (FITC) so as to facilitate microscopy observation. They were deposited by incubating a 200 μL liquid droplet for 45 min and finally rinsed thoroughly with PBS solution. Bacteria Deposition. E. coli DH5 alpha cells (exponential phase), chromosomally engineered to express green fluorescence (20) Francius, G.; Tesson, B.; Dague, E.; Martin-Jezequel, V.; Dufr^ene, Y. F. Environ. Microbiol. 2008, 10, 1344. (21) Dague, E.; Alsteens, D.; Latge, J.-P.; Dufrene, Y. Biophys. J. 2008, 94, 1. (22) Dague, E.; Alsteens, D.; Latge, J. P.; Verbelen, C.; Raze, D.; Baulard, A. R.; Dufr^ene, Y. F. Nano Lett. 2007, 7, 3026. (23) Alsteens, D.; Dague, E.; Rouxhet, P. G.; Baulard, A. R.; Dufrene, Y. F. Langmuir 2007, 23, 11977. (24) Alsteens, D.; Verbelen, C.; Dague, E.; Raze, D.; Baulard, A. R.; Dufr^ene, Y. F. Pflugers Arch. - Eur. J. Physiol. 2008, 456, 117. (25) Kienberger, F.; Rankl, C.; Pastushenko, V.; Zhu, R.; Blaas, D.; Hinterdorfer, P. Structure 2005, 13, 1247. (26) Hinterdorfer, P.; Dufr^ene, Y. F. Nat. Methods 2006, 3, 347. (27) Gad, M.; Ikai, A. Biophys. J. 1995, 69, 2226. (28) Cerf, A.; Cau, J.-C.; Vieu, C. Colloids Surf., B: Biointerfaces 2008, 65, 285.
5732
DOI: 10.1021/la9004642
Scheme 1. Schematic representation of the approach developed to generate a microstructured surface with functionalized patterns: (1) OTS μCP; (2) OTS μCP followed by epoxide or APTES solution incubation diluted to 2.5% (v/v) or 1% (v/v), respectively, with ethanol.
protein (GFP), were grown from a single colony in LuriaBertoni (LB) broth in a rotary shaker incubator at 37 °C overnight. These fluorescently labeled bacteria were centrifuged at 5000 g for 2 min. The supernatant was removed and bacteria were resuspended in PBS. A 300 μL liquid droplet was incubated for 20 min on the microstructured slide, and a coverslip was placed on top to obtain a monolayer of E. coli and enhance the efficiency of bacteria deposition. Then the coverslip was removed by gently pipetting a small amount of PBS to separate the coverslip from the substrate. Characterization Techniques. For quartz crystal microbalance analysis, we used an E4 Quartz Crystal Microbalance with Dissipation monitoring (QCM-D) from Q-Sense. Quartz surfaces were functionalized with the corresponding molecules using the same protocols as the ones described in the Surface Chemistry section. AFM images were performed using a BioscopeII (VEECO) in tapping and contact mode with tips purchased from VEECO (contact mode: Kc = 0.05 N/m, tapping mode Kc = 3 N/m). The AFM was equipped with a temperature controller to allow a thermal heating of the substrate. The experiments were conducted with temperatures ranging from 25 to 45 °C. For force spectroscopy (FS) experiments, the cantilevers spring constants were measured, for each lever, by the thermal tune method and were found to be around 0.05 N/m. The force curves were converted into indentation curves and fitted to the Hertz model taking into account a conical tip thanks to PUNIAS software.29
Results and Discussion Chemical Patterning. The chemical template was achieved using μCP30 in order to generate micrometric zones of preferential adsorption of bacteria. The antiadhesive layer of OTS containing micrometric uncoated zones was followed by a complementary surface functionalization of these bare regions with bacteria-binding capacity molecules. Our OTS μCP process reproduces the original stamp with high fidelity and with micrometer resolution, over a large area (cm2). We can obtain a homogeneous molecular layer of 4-5 nm thickness with an (29) Carl, P.; Kwok, C. H.; Manderson, G.; Speicher, D. W.; Disscher, D. E. Proc. Natl. Acad. Sci. U.S.A. 2001, 98, 1565. (30) Michel, B.; Bernard, A.; Bietsch, A.; Delamarche, E.; Geissler, M.; Juncker, D.; Kind, H.; Renault, J.-P.; Rothuizen, H.; Schmid, H.; Schmidt-Winkel, P.; Stutz, R.; Wolf, H. IBM J. Res. Dev. 2001, 45, 697.
Langmuir 2009, 25(10), 5731–5736
Cerf et al.
average roughness (root-mean-square (rms)) of 2 nm.28,31 The optimization of this step was crucial for the success of our approach. Then, to induce selectivity between untreated E. coli and the bare regions, we first evaluated through quartz crystal microbalance analysis the interaction between untreated bacteria and different candidate surface chemistries: OTS-coated and virgin SiO2 surfaces as references, and silanizated SiO2 surfaces with either a positively charged amino-silane (APTES) or an epoxide-terminated silane. The frequency variation between the initial state before untreated E. coli adsorption and the final state after rinsing and desorption of weakly bound cells, is used as an indication of the degree of the bacteria-surface interaction. This interaction, more precisely the mass of E. coli cells adsorbed, is 5 times stronger in the case of an electrostatic interaction between negatively charged E. coli and a positively charged surface (APTES), and is not significative in the case of an hydrophobic OTS-coated surface (Figure 1). It has to be noticed that quartz microbalance analysis only reveals the interaction between bacteria and the different plane functionalized surfaces, but does not take into account the effect of the surface microstructuration. The molecule finally selected to functionalize the patterns where we subsequently want to deposit untreated bacteria is the APTES silane. In order to chemically engineer the surface with strong adsorption contrast, we thus filled the uncoated OTS patterns obtained after μCP with APTES silane molecules through a simple incubation (Scheme 1). The microstructured OTS surface after APTES local functionalization was first characterized by AFM. The typical AFM image shown in Figure 2 illustrates the evolution of the surface topography after APTES silanization. Herein, by comparison with the OTS topographic profile described previously, we clearly see that the topographic relief is reduced; the grafting of APTES molecules selectively took place inside the patterns uncoated with OTS molecules. To further elucidate the exact location of APTES molecules and prove the accuracy of this local functionalization protocol, we used electrostatic bindings between APTES and negatively charged gold nanoparticles or colloids. According to Figure 3, effectively, colloids were clearly deposited inside the patterns (blue circles): 61 nanoparticles/μm2. The nanoparticle counting per square pixel area of the sample surface evaluated outside the patterns (10 nanoparticles/μm2) corresponds to the nonspecific deposition evaluated in a substrate without APTES. Thus, colloid incubation enabled us to colocalize APTES molecules. The results show that APTES molecules were correctly deposited on the SiO2 patterns. Bacteria Immobilization on Chemical Templates. After a 20 min incubation of a solution containing untreated bacteria over the microstructured surface with APTES functionalized patterns, dark field (Figure 4) and AFM images (Figure 5) revealed a selective attachment of untreated E. coli onto APTES patterns: the electrostatic binding we expected from our QCM analysis took place. Adjacent OTS patterns prevent the unspecific adsorption of bacteria on the surface. Figures 4 and 5 testify that we obtained here a very high success rate of individual adsorption events. Such uniformly distributed bacteria arrays covering an area of several millimeters were obtained repeatedly. Further from our expectations, additional experiments (not shown here) performed by incubation of the untreated bacteria onto a microstructured surface with SiO2 or epoxide patterns have shown that the selective adsorption (31) Cau, J.-C.; Cerf, A.; Thibault, C.; Genevieve, M.; Severac, C.; Peyrade, J.P.; Vieu, C. Microelectron. Eng. 2008, 85, 1143.
Langmuir 2009, 25(10), 5731–5736
Article
Figure 1. Comparison of adsorption and desorption kinetics of untreated E. coli bacteria onto SiO2 substrates (T = 25 °C) exhibiting different surface chemistries. Δf versus time for exposure of an OTS layer (yellow curve), an epoxide layer (blue curve), a SiO2 surface (green curve), and an APTES layer (red curve) to untreated E. coli.
Figure 2. OTS microstructured surface after incubation with APTES silane molecules. Height AFM image with the corresponding profile of a region of oval-shaped patterns (scale bar corresponds to 3 μm).
of bacteria onto these patterns also takes place with the same success rate, but with a weaker bounding. After vigorous rinsing of the substrates, we observe that bacteria are more loosely attached to SiO2 or epoxide patterns than to APTES patterns. In fact, the chemical micropatterning induces the trapping of bacteria in their natural media and mediates the dynamic assembly of bacteria into arranged arrays: the chemical template in itself is responsible for the selective adsorption process. The same kind of surface microstructuration can induce the trapping of biotinylated as well as nonbiotinylated bacteria. Hence, the additional local functionalization will only influence the binding strength between bacteria and the surface rather than their localization. This demonstrates that the surface microstructuration we have developed is really very DOI: 10.1021/la9004642
5733
Article
Cerf et al.
Figure 3. Colloid incubation to evidence APTES molecules deposition. Height AFM image of a region of the substrate with oval-shaped patterns (A) with the corresponding zoomed AFM height image (B). Scale bars correspond to 4 and 3 μm for A and B, respectively.
Figure 4. Untreated bacteria deposition on a microstructured surface with APTES functionalized patterns (1100 μm 1000 μm dark field image (left) with the corresponding zoomed image (right). Scale bar corresponds to 30 μm).
Figure 5. Untreated bacteria deposition on an OTS microstructured surface with APTES functionalized patterns. On the left, an AFM deflection image, and on the right, an AFM height image, both recorded in PBS.
efficient for producing highly ordered arrays of living bacteria with single-cell resolution, over a large area, without the need to treat the bacteria. A few examples of the results we can obtain by using different pattern geometries have been given as Supporting Information with this paper. We further performed cell viability tests using a fluorescence staining technique which proved that bacterial cells grafted on the chemically engineered 5734
DOI: 10.1021/la9004642
substrates are alive and that the thermal treatment damaged the cell membrane.28 Nanomechanical Properties of Alive or Dead Bacteria. Once we obtained a strong grafting of bacteria at registered positions, we could then conduct the FS studies on alive and dead bacteria. Figure 6 presents the same bacterium imaged by AFM in contact mode before (a,d) and after Langmuir 2009, 25(10), 5731–5736
Cerf et al.
Article
Figure 6. Mechanical properties of alive or dead bacteria: (a) AFM deflection image of single living E. coli bacterium. A foot print can be seen on the right of the cell. (d) AFM deflection image of the inset in panel a; (g) AFM deflection image of the same single bacteria killed by thermal treatment (20 min, 45 °C). The foot print has disappeared. (b,e,h) Elasticity maps (z-range = 10 MPa) corresponding, respectively, to images a, d, and g inset. (c,f,i) Elasticity distribution with a typical force curve corresponding to b, e, and h.
heating for 20 min at 45 °C (g). The cell is roughly 600 nm in height which is a classical value for gram negative bacteria imaged in contact mode32 but is not visually modified by the thermal treatment. In fact, neither the height nor the surface aspect is modified by such a treatment. Nevertheless, the foot print that can be seen in Figure 6a has disappeared in Figure 6g, suggesting that heating could desorb weakly attached substances. Figure 6b,e,f displays elasticity maps recorded on the living (b,e) and dead cell (f). The Young moduli are plotted as a function of a color scale ranging from 0 to 10 MPa. In other words, the brighter the pixel, the higher the Young modulus. To avoid any artifact due to edge effect described in the literature,33 force volumes were mostly recorded on a small zone (500 500 nm2) at the center of the cell (e,h) resulting in a homogeneous repartition of the Young moduli (f,i). In force volume b, the Young modulus of the cell and the foot print is 1.9 ( 0.9 MPa (r2> 0.98, n = 110), whereas it is 150 ( 13 MPa for the rest of the surface. In force volume e, which is recorded exclusively on the top of the cell, the Young modulus is 3.0 ( 0.6 MPa (r2 > 0.97, n = 1024). However, after heating (force volume h and histogram i) the elastic modulus increased to 6.1 ( 1.5 MPa (r2 > 0.93, n = 1024). Thus if it is impossible to distinguish (32) Gaboriaud, F.; Dague, E.; Bailet, S.; Duval, J. F.; Jorand, F.; Thomas, F. Colloids Surf., B: Biointerfaces 2006, 50, 123. (33) Gaboriaud, F.; Parcha, B. S.; Gee, M. L.; Holden, J. A.; Strugnell, R. A. Colloids Surf., B: Biointerfaces 2008, 62, 206.
Langmuir 2009, 25(10), 5731–5736
between a living or a dead cell by examinating AFM images, it is possible to do so by studying their mechanical properties. Recently, the Young modulus of Staphylococcus aureus cell wall has been determined before and after digestion by lysostaphin.34 E shows a dramatic decrease, from 1.8 ( 0.2 MPa before digestion, to 0.2 ( 0.05 MPa after the lysostaphin treatment. This is due to the cleavage of the peptidoglycan induced by the enzyme. In our experiments, E increases, which implies that the cell becomes stiffer after heating. E. coli is a Gram-negative bacterium, and its cell wall is constituted by an inner membrane, a periplasmic space, which contains a thin layer of peptidoglycan, and an outer membrane covered by lipopolysaccharides. E increase could be due to the collapse of the polysaccharide that recovers the bacteria. Moreover, the outer membrane is linked to the peptidoglycan by lipoproteins, namely the Braun lipoproteins, that are also amenable to fold under a different way after heating, resulting in an increase of the stiffness. Finally, one cannot rule out an increase of the turgor pressure after heating. Those hypotheses on the cell wall modification, which may explain the Young modulus increase, are summarized in Scheme 2. In conclusion, to probe the mechanical properties of living and dead bacteria via AFM indentation experimentations, we (34) Francius, G.; Domenech, O.; Mingeot-Leclercq, M. P.; Dufr^ene, Y. F. J. Bacteriol. 2008, 190, 7904.
DOI: 10.1021/la9004642
5735
Article
Cerf et al.
Scheme 2. Hypothesis on the action of the heating treatment on a Gram-negative bacteria cell wall. On the left, a simplified vision of a healthy cell wall is given, and on the right, the potential modifications are presented: collapse of the LPS, different folding of the Braun protein resulting in a decrease of the periplasmic space, increase of the cytoplasmic turgor pressure due to water afflux in the most concentrated compartment. Those three phenomena could explain the increase of the Young modulus.
first set up a fast, simple, and reproducible procedure enabling us to produce reliable chemical patterns exhibiting different surface properties to induce selective adsorption of individual bacteria in liquid media at registered positions. We have evidenced a selective adsorption of bacteria on these local chemical patterns, producing highly ordered arrays of single living bacteria with a success rate close to 100%. We then used this controlled immobilization method to study the mechanical properties of dead or alive bacterial cell in aqueous environment. We have demonstrated that the cells with a damaged membrane present a higher Young modulus (6.1 ( 1.5 MPa) than healthy cells (E=3.0 ( 0.6 MPa). At the same time, it has been impossible to evidence a difference between the AFM images of the living and the dead cell. Finally, the immobilization process also offers the possibility to tightly control the number and exact arrangement of bacteria deposition over a millimeter scale area, allowing us to draw bacteria patterns. By creating a pattern of surface structures or properties, our method involving soft lithography and chemical patterning can further be adapted to other applications. Further
5736
DOI: 10.1021/la9004642
extension of the technique could easily lead to generating arrays of any kind of microorganisms by simply using the appropriate pattern functionalization. Acknowledgment. The authors would like to thank Adilia Dagkessamanskaia, Olivier Guais, and Fabien Durand from the Biotechnology & Bioprocedures Laboratory (LBB Toulouse) for their help in providing fluorescent E. coli bacteria. They also wish to thank Mike Genevieve and the TEAM group from LAAS-CNRS for the design of the photolithography masks. We thank D. Alsteens for constructive discussion and our reviewers for helpful comments. E.D. is a researcher of Centre National de la Recherche Scientifique (CNRS). This work was supported by the EC-funded project NaPa (Contract No. NMP4-CT-2003-500120). Supporting Information Available: Fluorescence images of bacteria deposition on various patterned streptavidin substrates. This material is available free of charge via the Internet at http://pubs.acs.org.
Langmuir 2009, 25(10), 5731–5736
