pubs.acs.org/Langmuir © 2010 American Chemical Society
Use of Force Spectroscopy to Investigate the Adhesion of Living Adherent Cells Gilles Weder,*,†,‡ Nicolas Blondiaux,† Marta Giazzon,† Nadege Matthey,† Mona Klein,† Rapha€el Pugin,† Harry Heinzelmann,† and Martha Liley† † Swiss Centre for Electronics and Microtechnology, CSEM SA, Nanotechnology and Life Sciences, Jaquet-Droz 1, 2002 Neuch^ atel, Switzerland, and ‡Laboratory of Biosensors and Bioelectronics, Institute for Biomedical Engineering, ETH Zurich, Gloriastrasse 35, 8092 Z€ urich, Switzerland
Received December 1, 2009. Revised Manuscript Received January 27, 2010 The use of force spectroscopy to study the adhesion of living fibroblasts to their culture substrate was investigated. Both primary fibroblasts (PEMF) and a continuous cell line (3T3) were studied on quartz surfaces. Using a fibronectincoated AFM cantilever, it was possible to detach a large proportion of the 3T3 cells from the quartz surfaces. Their adhesion to the quartz surface and the effects of topography on this adhesion could be quantified. Three parameters characteristic of the adhesion were measured: the maximum force of detachment, the work of adhesion, and the distance of detachment. Few PEMF cells were detached under the same experimental conditions. The potential and limitations of this method in measuring cell/surface interactions for adherent cells are discussed.
Introduction Cell adhesion has been investigated using many techniques. Qualitative information about adhesion has been obtained using techniques based on microscopy. The use of fluorochrome-conjugated antibodies and green fluorescent protein (GFP) fusion proteins has produced high-resolution images of the distribution of binding proteins in cells.1-3 Fluorescence resonance energy transfer (FRET) has been used to determine the dynamic composition for protein-protein interactions.4 Alternatively, cells have been seeded on deformable substrates, and the adhesion forces have been correlated with the deformation of the underlying substrate.5 Semiquantitative results have been obtained by detaching cells using centrifugal force6 and hydrodynamic shear flow,7 which allows comparative cell adhesion measurements. Quantitative data has been obtained using microneedles8 to detach adherent cells. However, their resolution is limited and does not allow the investigation of cell adhesion in detail. In contrast, magnetic tweezers provide much greater sensitivity and have been used to probe unbinding forces during the initial stages of cellular adhesion.9 However, the maximal vertical force of 200 pN that can be applied greatly restricts their use in studying cell adhesion. One of the most powerful techniques used to quantify biological interactions is force spectroscopy. This AFM-based method is quantitative, has a very wide force range, and can be used in a *Corresponding author. E-mail:
[email protected]. (1) Cavalcanti-Adam, E. A.; Volberg, T.; Micoulet, A.; Kessler, H.; Geiger, B.; Spatz, J. P. Biophys. J. 2007, 92, 2964–2974. (2) Senesi, G. S.; D’Aloia, E.; Gristina, R.; Favia, P.; d’Agostino, R. Surf. Sci. 2007, 601, 1019–1025. (3) Tsien, R. Y. Annu. Rev. Biochem. 1998, 67, 509–544. (4) Kam, Z.; Volberg, T.; Geiger, B. J. Cell Sci. 1995, 108, 1051–1062. (5) Munevar, S.; Wang, Y.; Dembo, M. Biophys. J 2001, 80, 1744–1757. (6) Lotz, M. M.; Burdsal, C. A.; Erickson, H. P.; McClay, D. R. J. Cell Biol. 1989, 109, 1795–1805. (7) Alon, R.; Hammer, D. A.; Springer, T. A. Nature 1995, 374, 539–542. (8) Athanasiou, K. A.; Thoma, B. S.; Lanctot, D. R.; Shin, D.; Agrawal, C. M.; LeBaron, R. G. Biomaterials 1999, 20, 2405–2415. (9) Walter, N.; Selhuber, C.; Kessler, H.; Spatz, J. P. Nano Lett. 2006, 6, 398– 402. (10) Micoulet, A.; Spatz, J. P.; Ott, A. ChemPhysChem 2005, 6, 663–670.
8180 DOI: 10.1021/la904526u
number of commercially available instruments.10 It has been widely used since the mid-1990s to study biomolecular interactions.11-13 More recently, a number of different experimental approaches to quantifying cell adhesion using force spectroscopy have been reported. Most of these studies focus on individual membrane proteins and their interactions with a surface when the cell is brought briefly into contact with it. In these reports, single trypsinized cells were first captured on a functionalized AFM cantilever by gently pressing it onto the cell. This converted the living cell into a probe that could be brought into well-defined contact with a chosen surface. In this configuration, cell adhesion has been investigated either at the cell level14,15 or at the singlemolecule level16-18 depending on the proteins expressed at the cell surface, the surface decoration, and the contact time, typically between a few seconds and a few minutes. In contrast, there are very few studies on the use of force spectroscopy to investigate cell/surface interactions in adherent cells in situ (i.e., without trypsinisation to remove them from their culture substrate). Those studies that exist measured the forces necessary to detach dead and chemically fixed cells from surfaces. Whereas this approach may give some information on cellular adhesion in the living cell, this information is clearly qualitative and is vulnerable to the presence of artifacts induced by fixing. In this article, we investigate the use of force spectroscopy to directly measure the adhesion of living adherence-dependent cells to their culture substrates. We have studied the influence of two parameters on cell adhesion: surface topography and cell type. (11) Florin, E. L.; Moy, V. T.; Gaub, H. E. Science 1994, 264, 415–417. (12) Krautbauer, R.; Rief, M.; Gaub, H. E. Nano Lett. 2003, 3, 493–496. (13) Rief, M.; Gautel, M.; Oesterhelt, F.; Fernandez, J. M.; Gaub, H. E. Science 1997, 276, 1109–1112. (14) Puech, P. H.; Taubenberger, A.; Ulrich, F.; Krieg, M.; Muller, D. J.; Heisenberg, C. P. J. Cell Sci. 2005, 118, 4199–4206. (15) Weder, G.; V€or€os, J.; Giazzon, M.; Matthey, N.; Heinzelmann, H.; Liley, M. Biointerphases 2009, 4, 27–34. (16) Friedrichs, J.; Torkko, J. M.; Helenius, J.; Teravainen, T. P.; Fullekrug, J.; Muller, D. J.; Simons, K.; Manninen, A. J. Biol. Chem. 2007, 282, 29375–29383. (17) Taubenberger, A.; Cisneros, D. A.; Friedrichs, J.; Puech, P. H.; Muller, D. J.; Franz, C. M. Mol. Biol. Cell 2007, 18, 1634–1644. (18) Zhang, X.; Wojcikiewicz, E.; Moy, V. T. Biophys. J. 2002, 83, 2270–2279.
Published on Web 02/10/2010
Langmuir 2010, 26(11), 8180–8186
Weder et al.
Article
Adherence-dependent cells are profoundly influenced by the topography of the substrate to which they adhere. Cellular responses such as adhesion, morphology, alignment, guidance, gene expression, and proliferation are all influenced by topography.19-24 The effects may differ widely depending on the dimensions and form of the topography25 and the cell type.26 A large and sometimes contradictory body of literature on this topic indicates the complexity of cell response to topography. However, for the example chosen in this study;fibroblasts and submicrometer topographies;most authors report inhibition of adhesion and resulting cell rounding. To quantify cell/surface interactions, living adherent cells were detached from their substrates by pulling vertically with an AFM cantilever. During the detachment process, the deflection of the cantilever, and thus the force applied to the cell, was constantly recorded.27-29 Cell adhesion was quantified using three parameters: the maximum force applied during the detachment process, the work of detachment, and the distance of detachment. Four different examples of cell adhesion were studied. The adhesion of 3T3 mouse fibroblasts to polylysine-treated quartz surfaces after 24 h of growth in situ was first quantified. It was then compared with the adhesion of 3T3 cells on identical quartz surfaces that had been structured with two different submicrometer pillar motifs. Finally, the 3T3 cells were replaced with primary mouse fibroblasts, and a similar study was carried out on flat polylysine-treated quartz.
Materials and Methods Microfabrication of Low-Aspect-Ratio Pillars in Quartz. Low-aspect-ratio (270-nm-high) quartz pillars were fabricated using a modified procedure previously reported for silicon.30 Four inch quartz wafers (Si-Mat, Germany) were cleaned in piranha solution (H2SO4/H2O2 4: 1 v/v for 10 min at 120 °C) and rinsed in flowing water (Milli-Q 185 plus, Millipore AG, Switzerland). Caution! Piranha solution reacts violently with all organics and should be handled with care. The substrates were dried in a flow of nitrogen under a laminar flow hood just before use. In the first step, a polymer blend thin film was deposited on the surface. Poly(methyl methacrylate) (PMMA, Mw = 106 kDa, Polymer Standard Services, Germany) and polystyrene (PS, Mw = 101 kDa, Polymer Standard Services, Germany) were dissolved in dioxane to a concentration of 15 mg/mL to give stock solutions. These were mixed to obtain a PS/PMMA ratio of 30:70 w/w. The polymer blend solution was then spin coated onto the substrate under a controlled atmosphere (T = 21 °C and RH = 35%) using a two-step spin-coating procedure: 2 s at 500 rpm followed by 60 s (19) Charest, J. L.; Garcιa, A. J.; King, W. P. Biomaterials 2007, 28, 2202–2210. (20) Dalby, M. J.; Riehle, M. O.; Yarwood, S. J.; Wilkinson, C. D. W.; Curtis, A. S. G. Exp. Cell Res. 2003, 284, 272–280. (21) Fitton, J. H.; Dalton, B. A.; Beumer, G.; Johnson, G.; Griesser, H. J.; Steele, J. G. J. Biomed. Mater. Res. 1998, 42, 245–257. (22) Rajnicek, A.; McCaig, C. J. Cell Sci. 1997, 110, 2915–2924. (23) Wan, Y.; Wang, Y.; Liu, Z.; Qu, X.; Han, B.; Bei, J.; Wang, S. Biomaterials 2005, 26, 4453–4459. (24) Zinger, O.; Anselme, K.; Denzer, A.; Habersetzer, P.; Wieland, M.; Jeanfils, J.; Hardouin, P.; Landolt, D. Biomaterials 2004, 25, 2695–2711. (25) Dalby, M. J.; Riehle, M. O.; Johnstone, H. J. H.; Affrossman, S.; Curtis, A. S. G. Tissue Eng. 2002, 8, 1099–1108. (26) Dalby, M. J.; Marshall, G. E.; Johnstone, H. J. H.; Affrossman, S.; Riehle, M. O. IEEE Trans. Nanobiosci. 2002, 1, 18–23. (27) Franz, C. M.; Taubenberger, A.; Puech, P. H.; Muller, D. J. Sci. STKE 2007, 2007, pl5–pl5. (28) Helenius, J.; Heisenberg, C. P.; Gaub, H. E.; Muller, D. J. J. Cell Sci. 2008, 121, 1785–1791. (29) Wojcikiewicz, E. P.; Zhang, X.; Chen, A.; Moy, V. T. J. Cell Sci. 2003, 116, 2531–2539. (30) Blondiaux, N.; Scolan, E.; Popa, A. M.; Gavillet, J.; Pugin, R. Appl. Surf. Sci. 2009, 256, S46–S53.
Langmuir 2010, 26(11), 8180–8186
at 2000 rpm. The resulting thin film was structured into two phases, one of which (PS) was removed by rinsing the sample in cyclohexane to reveal the quartz surface underlying the PS areas. In the second step, a durable etch mask was fabricated from the thin polymer film. The polymer-coated surface was first exposed to oxygen plasma (O2, 50 sccm, 0.05 Torr, 15 W) for 90 s using RIE plasma (Plasmalab 80plus, Oxford Instruments, U.K.) to remove any residual PS layer and expose the quartz substrate. A 10-nm-thick chromium layer was then deposited on the sample using a thermal evaporator (Lab 600H, Leybold Optics, Germany). The remaining PMMA phase and its chromium layer were removed by sonicating the sample in acetone. The last step was the etching of the quartz, which was carried out in a deep reactive ion etcher (DRIE) (AMS 200, Alcatel, France) using fluorine chemistry. After being etched, the samples were dipped in chromium etchant CR-7 for 5 min, then cleaned in SC-1 solution (H2O/NH4OH/H2O2 5:1:1 v/v for 10 min at 70 °C, and finally exposed to oxygen plasma (RIE, O2, 100 sccm, 0.2 Torr, 100 W) for 1 h.
Microfabrication of High-Aspect-Ratio Pillars in Quartz. A 200 nm layer of amorphous silicon (a-Si) was first deposited onto quartz by means of LPCVD (Centrotherm Furnace, Germany). The chromium etch mask was then fabricated on top of the a-Si layer as described above. The chromium layer was used as an intermediate etch mask that was transferred onto the a-Si layer using DRIE. The good etch selectivity and the relatively thick (200 nm) a-Si mask allowed the fabrication of deeper structures in the quartz. After fabrication, the cleaning procedure was the same as for 270 nm pillars except that the samples were also exposed to KOH after the chromium etchant step in order to remove the a-Si layer. Prior to cell seeding, the quartz surfaces were activated with oxygen plasma (Harrick Plasma, NY) for 10 min and then incubated in a solution of 0.01% poly-L-lysine (MW = 150300 kDa, Sigma, MO) for 30 min to promote cell attachment. Cell Culture. Mouse fibroblast cell line 3T3 was obtained from Marinpharm GmbH (Luckenwalde, Germany) and was maintained in a continuous culture in Dulbecco’s Modified Eagle’s Medium (DMEM) supplemented with 10% heat-inactivated standardized fetal bovine serum (FBS, Biochrom AG, Germany), 50 units/mL of penicillin (Sigma, MO), 50 μg/mL of streptomycin (Sigma, MO), 1.5 mM of L-glutamine (Sigma, MO), and 1% nonessential amino acids (Bioconcept, Switzerland) at 37 °C in a humidified 5% CO2 atmosphere. Primary mouse embryonic fibroblasts (PMEF) were isolated from embryos derived from C57 mothers between 12 and 14 days of gestation. Embryos were cut into small pieces and incubated in a solution of 2.5 g/L trypsin and 0.38 g/L ethylenediaminetetraacetic acid (EDTA) for 20 min at 37 °C under agitation after removing the extracellular tissue, head, and liver. Trypsin was inactivated by the addition of DMEM containing 10% FCS, and cells were purified by successive steps of filtration and centrifugation. Cells were maintained in DMEM supplemented with 10% FBS, 50 units/mL penicillin, and 50 μg/mL streptomycin at 37 °C in a humidified 5% CO2 atmosphere. Cell Seeding. Cells were detached from the culture dishes by incubating with a solution of 2.5 g/L trypsin and 0.38 g/L of EDTA for 5 min. The dissociated cells were seeded following the regular conditions of cultivation on the modified quartz surfaces and were incubated for 24 h to allow the cells to spread and their adhesion proteins to be renewed. Cell detachment experiments were performed in regular medium supplemented with 25 mM 4-(2-hydroxaethyl)-1-piperazineethanesulfonic acid (HEPES) at 37 °C in a temperature-controlled chamber. Atomic Force Microscopy. We used a Nanowizard II atomic force microscope (JPK Instruments, Germany) mounted on an inverted optical microscope (Carl Zeiss, Germany). A CellHesion module (JPK Instruments, Germany) allowed vertical displacements of the force microscope cantilever of up to 100 μm, and an DOI: 10.1021/la904526u
8181
Article
Weder et al.
Figure 1. (A) Schematic of the detachment of a single adherent cell by force spectroscopy and (B) a corresponding real force-distance curve. In the initial phase of the approach, there is no contact between the cantilever and the adherent cell (A I-II). The cantilever is pressed onto the adherent cell until a preset maximal force of 1.5 nN is attained (A III). The position of the cantilever is held constant for 30 min before it is withdrawn at 0.5 μm/s (A IV). Finally, the cell is either on the cantilever (A V, possibility 1) or on the culture surface (A VI, possibility 2). During the separation of the cell from the surface (or from the cantilever), a force-distance curve is recorded (B), indicating the maximum force and distance of detachment. The hatched area shows the work of detachment. In both panels, green defines the approach and violet defines the retraction.
incubation chamber (JPK Instruments, Germany) maintained the sample at the desired temperature. Silicon tipless cantilevers (NSC12, Mikromasch, Estonia) that were 130 μm long with a rectangular shape and a nominal spring constant of 4.5 N/m were used for all force spectroscopy AFM measurements. Cantilevers were calibrated in water before each measurement using the thermal fluctuation method.31 AFM cantilevers were functionalized with fibronectin to promote cell adhesion.32 They were cleaned in piranha solution at 120 °C for 10 min and activated with oxygen plasma (Harrick Plasma, NY) for 3 min. The cantilevers were then incubated in a solution containing 94% acidic methanol (1 mM acetic acid in methanol), 5% water, and 1% (3-aminopropyl)-triethoxysilane (Sigma, MO) for 30 min. The cantilevers were extensively rinsed in methanol and dried before incubation in a solution of 1% glutaraldehyde for 1 min. They were finally rinsed in water and incubated in a solution of 10 μg/mL fibronectin (Sigma, MO) for 1 h. The cantilevers were kept in the solution of fibronectin until use. Force-Distance Curves. The extremity of a fibronectindecorated cantilever was precisely positioned above the targeted cell. Representative cells of comparable size were selected for each experimental condition. The cantilever approached the cell until a repulsive force of 1.5 nN was reached. It was left in contact with the cell at constant height for 30 min before the retraction step was started. Force-distance curves were acquired with an approaching and retracting speed of 0.5 μm/s using the maximum z range of 100 μm. A new cell sample and a new cantilever were used for each measurement. Data Processing and Statistical Analysis. The mechanical parameters related to cell detachment;the maximum force, work, and distance of detachment;were obtained from the force-distance curves (Figure 1B). The work of detachment was obtained by integrating the adhesion forces over the distance traveled by the cantilever. Data analysis was carried out using Image Processing software (JPK Instruments, Germany). For each condition, 1 force dis(31) Hutter, J. L.; Bechhoefer, J. Rev. Sci. Instrum. 1993, 64, 1868–1873. (32) Selhuber, C., Ph.D. Thesis, Heidelberg University, Germany, 2006, pp 1-150.
8182 DOI: 10.1021/la904526u
tance curve was obtained from each of 20 different cells, giving 20 measurements. The mean of these values together with the standard error was then taken to be characteristic of each condition. The percentage of cells that detached from the surface was determined by optically examining the end of the AFM cantilever after each measurement. Values for different experimental conditions were compared using statistical analysis. Logarithmic transformations giving normal distributions of the mechanical parameters were analyzed using the two-sample t test. The Bonferroni correction was used for multiple comparisons to account for the inflation of the type I error. For all statistical tests, a p value
