Simultaneous Imaging and Chemical Attack of a Single Living Cell

Dec 7, 2010 - Single Living Cell within a Confluent Cell. Monolayer by Means of Scanning Electrochemical. Microscopy. Stefan Bergner, Joachim Wegener,...
0 downloads 0 Views 2MB Size
Download PDF
Anal. Chem. 2011, 83, 169–174

Simultaneous Imaging and Chemical Attack of a Single Living Cell within a Confluent Cell Monolayer by Means of Scanning Electrochemical Microscopy Stefan Bergner, Joachim Wegener, and Frank-Michael Matysik* Institute of Analytical Chemistry, Chemo- und Biosensors, University of Regensburg, 93053 Regensburg, Germany Epithelial cell monolayers from rat kidney were imaged by scanning electrochemical microscopy (SECM) with sub-micrometer resolution in both lateral and vertical direction. Platinum disk ultra-microelectrodes (UMEs) with effective electrode radii between 200 and 600 nm were operated in the constant-height mode. The quality of the recorded SECM images compare favorably with those of phase contrast and confocal laser scanning microscopy. Besides the acquisition of SECM images, the UME was used to selectively attack a single living cell within the monolayer ensemble. Hydroxide ions were locally generated in the vicinity of a single target cell by the UME. The increase in pH induced cell necrosis that was subsequently imaged by SECM. It could be clearly demonstrated that the single target cell was selectively affected, whereas the adjacent reference cells remained unchanged. Scanning electrochemical microscopy (SECM) has been demonstrated to be a powerful tool to investigate morphological and chemical characteristics of microstructured substrates since its introduction in the late 1980s by Bard and co-workers.1 The use of SECM opens a wide variety of applications. Early investigations included imaging of inorganic microstructures2 and of microstructuring metal substrates.3 Because of its noninvasive nature, SECM is especially advantageous to study biological systems. However, when biological samples are investigated with ultra-microelectrodes (UMEs), care has to be taken to avoid adsorption on the electrode, mechanical tip-surface interactions, or even tip crashes. Important biological applications of SECM involve the oxidation or reduction of enzymatically generated compounds. Typically, the enzyme is immobilized on a support, and the product of the enzyme reaction is electrochemically detected.4 Further work involved the investigation of membrane transport mediated by protein channels5 that are crucial for different cellular functions. Only a few months after the advent of the SECM technique, the * Corresponding author. Tel: +49 0941/9434548. E-mail: frank-michael.matysik@ chemie.uni-regensburg.de. (1) Bard, A. J.; Fan, F. R. F.; Kwak, J.; Lev, O. Anal. Chem. 1989, 61, 132– 138. (2) Kwak, J.; Bard, A. J. Anal. Chem. 1989, 61, 1794–1799. (3) Husser, O. E.; Craston, D. H.; Bard, A. J. J. Electrochem. Soc. 1989, 136, 3222–3229. (4) Pierce, D. T.; Bard, A. J. Anal. Chem. 1993, 65, 3598–3604. 10.1021/ac1021375  2011 American Chemical Society Published on Web 12/07/2010

potential to image the stomata in living plant cells was recognized.6 Topographic SECM images of living cells are, however, rarely described in the literature.7-11 Over the last decades, several research initiatives have addressed the cellular release of chemical substances, such as oxygen during photosynthesis7 or even insulin.12 Since cellular viability can be determined by their respiratory activity, the oxygen concentration profile in the vicinity of single cells7 and the toxicity of substances such as cyanide13 have been studied. In a recent publication quantitative viability measurements of single living cells were reported. The respiratory activity was compared before and after the addition of a millimolar concentration of potassium cyanide.14 Further SECM studies have characterized the individual redox activity of cells in order to distinguish between metastatic and nontransformed human breast cells.15 Typically, topographic SECM images of single cells cannot compete with those acquired by common light microscopic techniques in live cell imaging, such as phase contrast or confocal laser scanning microscopy (CLSM). Since most of these images have been acquired using UMEs with electrode diameters in the range of 10-25 µm, they cannot provide sub-micrometer resolution in principle. Many efforts have been made to improve the quality of SECM images by incorporating various devices providing independent distance controls such as atomic16-21 or shear (5) Mauzeroll, J.; Bard, A. J. Proc. Natl. Acad. Sci. U.S.A. 2004, 101, 7862– 7867. (6) Lee, C.; Kwak, J. Y.; Bard, A. J. Proc. Natl. Acad. Sci. U.S.A. 1990, 87, 1740–1743. (7) Yasukawa, T.; Kondo, Y.; Matsue, T. Chem. Lett. 1998, 8, 767–768. (8) Yasukawa, T.; Kaya, T.; Matsue, T. Electroanalysis 2000, 12, 653–659. (9) Takahashi, Y.; Miyamoto, T.; Shiku, H.; Asano, R.; Yasukawa, T.; Kumagai, I.; Matsue, T. Anal. Chem. 2009, 81, 2785–2790. (10) Liu, B.; Cheng, W.; Rotenberg, A.; Mirkin, M. V. J. Electroanal. Chem. 2001, 500, 590–597. (11) Li, X.; Bard, A. J. J. Electroanal. Chem. 2009, 628, 35–42. (12) Kennedy, R. T.; Huang, L.; Atkinson, M. A.; Dush, P. Anal. Chem. 1993, 65, 1882–1887. (13) Kaya, T.; Torisawa, Y.; Oyamatsu, D.; Nishizawa, M.; Matsue, T. Biosens. Bioelectron. 2003, 18, 1379–1383. (14) Zhu, L.; Gao, N.; Zhang, X.; Jin, W. Talanta 2008, 77, 804–808. (15) Liu, B.; Rotenberg, S.; Mirkin, M. Proc. Natl. Acad. Sci. 2000, 97, 9855– 9860. (16) Macpherson, J. V.; Unwin, P. R. Anal. Chem. 2000, 72, 276–285. (17) Burt, D. P.; Wilson, N. R.; Weaver, J. M. R.; Dobson, P. S.; Macpherson, J. P. Nano Lett. 2005, 5, 639–643. (18) Dobson, P. S.; Weaver, J. M. R.; Burt, D. P.; Holder, M. N.; Wilson, N. R.; Unwin, P. R.; Macpherson, J. P. Phys. Chem. Chem. Phys. 2006, 8, 3909– 3914.

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

169

force microscopy,22-24 or have been based on the impedimetric signal25,26 or faradaic current.27,28 For example, recording the amperometric signal of the mediator oxidation or reduction is advantageous for topographical imaging because it allows a simple, reliable, and inexpensive distance control. The lateral resolution in SECM can be improved by using UMEs with sub-micrometer electrode radii, so-called nanoelectrodes or nanodes. Additionally, Heinze and co-workers29 reported the concept of the chemical lens to enhance the resolution by focusing the diffusion field. Baur and co-workers used a 1 µm carbon-ring UME under constantcurrent distance control to record a high-quality micrograph of an isolated PC12 cell and its neurites.28 In a recent report Mirkin et al. presented the detailed surface topography of a human breast epithelial cell using a 123 nm Pt probe in conjunction with constant-current SECM.30 Besides the use of UMEs as scanning probes, nanometer-sized optical fiber electrodes have attracted growing interest from several research groups. When these probes were used, a PC12 cell neurite,31 PC12,32 and HeLa cells33 were imaged in high resolution. In very recent developments the combination of scanning ion conductance microscopy (SICM)34 and SECM or intermittent contact-SECM35 was demonstrated, providing complementary topographical and electrochemical information. On account of their small dimensions, UMEs can be used to treat a substrate in high spatial resolution. Self-assembled monolayers36 and polymers37 were microstructured using UMEs. Enzymes were subsequently linked to the microstructured substrates, and the reaction products were detected in the substrategeneration tip-collection or the feedback mode. In addition, patterning of living cells has been extensively attempted by (19) Gullo, M. R.; Frederix, P. L. T. M.; Akiyama, T.; Engel, A.; deRooij, N. F.; Staufer, U. Anal. Chem. 2006, 78, 5436–5442. (20) Frederix, P. L. T. M.; Bosshart, P. D.; Akiyama, T.; Chami, M.; Gullo, M. R.; Blackstock, J. J.; Dooleweerdt, K.; de Rooij, N. F.; Staufer, U.; Engel, A. Nanotechnology 2008, 19, 384004/1–384004/10. (21) Anne, A.; Cambril, E.; Chovin, A.; Demaille, C.; Goyer, C. ACS Nano 2009, 3, 2927–2940. (22) Hengstenberg, A.; Blochl, A.; Dietzel, I. D.; Schuhmann, W. Angew. Chem., Int. Ed. 2001, 40, 905–908. (23) Katemann, B. B.; Schulte, A.; Schuhmann, W. Electroanalysis 2004, 16, 60–65. (24) Takahashi, Y.; Shiku, H.; Murata, T.; Yasukawa, T.; Matsue, T. Anal. Chem. 2009, 81, 9674–9681. (25) Alpuche-Aviles, M. A.; Wipf, D. O. Anal. Chem. 2001, 73, 4873–4881. (26) Ervin, E. N.; White, H. S.; Baker, L. A. Anal. Chem. 2005, 77, 5564–5569. (27) Isik, S.; Etienne, M.; Oni, J.; Blochl, A.; Reiter, S.; Schuhmann, W. Anal. Chem. 2004, 76, 6389–6394. (28) Kurulugama, R. T.; Wipf, D. O.; Takacs, S. O.; Pongmayteegul, S.; Garris, P. A.; Baur, J. E. Anal. Chem. 2005, 77, 1111–1117. (29) Borgwarth, K.; Heinze, J. J. Electrochem. Soc. 1999, 146, 3285–3289. (30) Sun, P.; Laforge, F. O.; Abeyweera, T. P.; Rotenberg, S. A.; Carpino, J.; Mirkin, M. V. Proc. Natl. Acad. Sci. U.S.A. 2008, 105, 443–448. (31) Maruyama, K.; Ohkawa, H.; Ogawa, S.; Ueda, A.; Niwa, O.; Suzuki, K. Anal. Chem. 2006, 78, 1904–1912. (32) Takahashi, Y.; Hirano, Y.; Yasukawa, T.; Shiku, H. Langmuir 2006, 22, 10299–10306. (33) Takahashi, Y.; Shiku, H.; Murata, T.; Yasukawa, T. Anal. Chem. 2009, 81, 9674–9681. (34) Takahashi, Y.; Shevchuk, A. I.; Novak, P.; Murakami, Y.; Shiku, H.; Korchev, Y. E.; Matsue, T. J. Am. Chem. Soc. 2010, 132, 10118–10126. (35) McKelvey, K.; Edwards, M. A.; Unwin, P. A. Anal. Chem. 2010, 82, 6334– 6337. (36) Wilhelm, T.; Wittstock, G. Electrochim. Acta 2001, 47, 275–281. (37) Kranz, C.; Wittstock, G.; Wohlschla¨ger, H.; Schuhmann, W. Electrochim. Acta 1997, 42, 3105–3111.

170

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

incorporating microcontact printing.38 Using this technique, growth regions of HeLa cells on glass substrates were defined.39 The respiratory activity of the cells was subsequently imaged by SECM. In recent reports the direct manipulation of immobilized enzymes by in situ generation of bromine40 and hydroxide ions41 using nanometer-sized UMEs were presented. In this work it is demonstrated that epithelial cell monolayers are very suitable substrates for SECM investigations based on the constant-height mode. We present SECM images of living normal rat kidney (NRK) cell monolayers with sub-micrometer resolution. Pt nanoelectrodes were used for both imaging and treating a single living cell chemically within the monolayer ensemble. EXPERIMENTAL SECTION Electrodes. Platinum UMEs were prepared using 25 µm platinum wires (Goodfellow, UK). The microwire was soldered to a copper lead. The etching process was performed in a drop of a solution consisting of saturated CaCl2 (60% v/v), H2O (36% v/v), and concentrated HCl (4% v/v)42 using differential pulse amperometry with potential pulses of ±2 V at a frequency of 50 Hz. The electrochemically sharpened Pt wire was subsequently sealed in a drawn-out soda-lime glass capillary. A diskshaped electrode was exposed by polishing with alumina polishing foils of decreasing grain size (30, 10, 3, 0.3 µm). A conically sharpened glass tip was consecutively prepared under a microscope using a converted hard disk with attached 0.3 µm alumina polishing foil. A thermal treatment was used to improve the quality of sealing. For electrical heating, a Kanthal wire (20 cm, 0.4 mm) was coiled in 10 loops and was heated by a DC power supply (10 V) for 5-10 s. The quality of the UMEs was characterized by steady-state cyclic voltammetry and probe approach curves over glass slides using a solution containing 1.5 mM ferrocene methanol and 0.25 M KNO3. Only those electrodes exhibiting a behavior in agreement with theory were used for further studies. A three-electrode arrangement with a Ag/AgCl/1 M KCl reference electrode (CH Instruments, Austin, TX) and a Pt auxiliary electrode was used throughout all experiments of this study. Equipment. A commercial SECM system was used for all experiments (CHI 920c, CH Instruments). The measurements were carried out in a homemade electrochemical cell (Teflon) with two clamps (PEEK) and four integrated screws (Teflon) for substrate fixation. The cell was bolted onto a stainless steel carrier that could be leveled. Confocal laser scanning microscopy was performed using a NIKON C1 multicolor confocal microscope together with a 60-fold water immersion objective (NA ) 0.9). For excitation, a laser line at 488 nm was used. Cell Culture. Normal rat kidney epithelial cells (NRK-52E) were obtained from the German Collection of Microorganisms and Cell Cultures (DSMZ; accession number ACC 199). They were cultured on 0.17 mm thick glass coverslips inside a 10 cm2 (38) Xia, Y.; Whitesides, G. M. Angew. Chem., Int. Ed. 1998, 37, 550–575. (39) Nishizawa, M.; Takoh, K.; Matsue, T. Langmuir 2002, 18, 3645–3649. (40) Li, X.; Geng, Q.; Wang, Y.; Si, Z.; Jiang, W.; Zhang, X.; Jin, W. Electrochim. Acta 2007, 53, 2016–2024. (41) Li, X.; Geng, Q.; Wang, Y.; Si, Z.; Jiang, W.; Zhang, X.; Jin, W. Electroanalysis 2007, 19, 1734–1740. (42) Lee, C.; Miller, C. J.; Bard, A. J. Anal. Chem. 1991, 63, 78–83.

Figure 1. Phase contrast micrograph of a confluent NRK cell monolayer cultured on a glass coverslip.

Petri dish. The NRK medium consisted of 5% fetal calf serum, 2 mM L-glutamine, and 100 µg/mL penicillin/streptomycin, respectively. Experimental buffers were based on Dulbecco’s phosphate-buffered saline solution (Dulbecco’s MEM, SigmaAldrich) and contained 136.9 mM NaCl, 2.7 mM KCl, 8.1 mM Na2HPO4, 1.5 mM KH2PO4, 0.5 mM MgCl2 · 6H2O, and 1.0 mM CaCl2 · 2H2O (PBS++). The cultures were kept in a humidified cell incubator at 37 °C containing 5% CO2 in the atmosphere. The growth medium was exchanged every 2-3 days. It took about 1 week until a confluent cell monolayer was formed from 1:20 dilution during routine subculture, which was controlled using a phase contrast microscope. Prior to SECM experiments, the growth medium was removed and the cells were washed five times with PBS++ buffer, consisting of 1 g/L D-glucose, until all floating cells were removed. Chemicals. The following substances were used during the mediator evaluation: [Ru(NH3)6]Cl3 (98%, Sigma, Steinheim, Germany), ferrocene methanol (FcMeOH, 99%, ABCR, Karlsruhe, Germany), K4[Fe(CN)6] · 3H2O (g98.5%, Sigma-Aldrich, Steinheim, Germany), and ferrocene carboxylic acid (FcCOOH, 98%, Alfa Aesar, Karlsruhe, Germany). K4[W(CN)8] · 2H2O and K4[Os(CN)6] · 3H2O were prepared according to the literature.43,44

Membrane-impermeable FITC-dextran (2 mg/mL; 464 kDa, Sigma-Aldrich) was used to label the extracellular space with bright fluorescence which allows the cell volume changes to be monitored with confocal laser scanning microscopy. All other chemicals were of analytical reagent grade. Solutions. For SECM and CLSM imaging, 1.5 mM FcMeOH in PBS++ buffer containing 1 g/L D-glucose was used. For the latter, FITC-dextran (2 mg/mL) was subsequently added. A PBS++ buffer with a reduced buffer capacity (one-tenth of the above standard PBS++) was used for the studies concerning local generation of hydroxide ions. A 0.25 M NaCl solution was used to adjust the osmolarity of the measuring solutions to 320 mosmol/kg, which was determined using a cryoscopic osmometer (Osmomat 030, Gonotec GmbH, Germany). All solutions were prepared in ultrapure water with a resistivity greater than 18 MΩ · cm (membraPure, Bodenheim). RESULTS AND DISCUSSION Morphology of NRK Cell Monolayers. Epithelial cell layers cover all inner and outer surfaces of mammalian organisms. They develop a typical cobblestone morphology with cell diameters in the range of about 10-30 µm when they are grown to confluence. Figure 1 shows a representative phase contrast micrograph of an NRK cell monolayer. In phase contrast microscopy optically denser parts of the cell bodies appear dark, whereas cell-cell contacts are rather bright. Studying NRK Cell Necrosis by Confocal Laser Scanning Microscopy (CLSM). The cell response to an increase in pH was investigated by CLSM. Cells were grown to confluence and were visualized by adding membrane-impermeable FITCdextran to the extracellular fluid which renders the extracellular space brightly fluorescent whereas the cell bodies remain dark. This labeling of the bulk phase avoids any problems typically arising from fluorophore bleaching. During image processing, the contrast was inverted such that the cell interior was bright while the extracellular space remained dark. In Figure 2A a confluent NRK cell monolayer is shown at pH 7.2 in optical xy and xz sections. The top of the individual cells are only a little higher in the z direction than the cell-cell contacts which are known to be

Figure 2. Images of xy (upper part) and xz (lower part) sections of confluent NRK monolayers recorded with a confocal laser scanning microscope. Pictures (A-C) demonstrate the effect of the pH (7.2, 8.0, and 10.0) on the morphology of the monolayer. Image contrast was achieved by adding FITC-dextran to the extracellular buffer and contrast inversion during image processing. Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

171

Figure 3. Approach curves of a 0.6 µm radius UME at three different positions (1-3) over the NRK cell monolayer. The curves were recorded from a fixed plane z0 approaching the cell monolayer. Because the point of the closest approach is not uniform, the travel distance is indicated relative to the parameter z0. The inset shows the corresponding positions within the line scan.

located at the apical pole between adjacent cells. The height differences between the highest point of the cell bodies and the coalescence point of cell junctions were calculated from the corresponding xz sections and were about (2.5 ± 0.5) µm on average. By adding 90 µL of NaOH (10 mM) to the bathing fluid, the pH value was subsequently increased to a pH of 8.0. The solution was allowed to stand for 5 min to ensure a homogeneous distribution. Figure 2B illustrates the cell monolayer after the addition of NaOH. The change in contrast is due to the pH dependent fluorescence emission of FITC. The increased pH affects the height of the apical cell surface, whereas the z position of the cell-cell contacts remains nearly unaffected. The height differences between the uppermost point of the cell bodies and coalescence point of the cell-cell junctions were slightly increased to an average of (3.0 ± 0.3) µm. After the subsequent addition of another 10 µL of NaOH, the pH increased to 10.0. It can be seen in Figure 2C that the cells are significantly swollen. The corresponding optical xy section images the upper part of the cell body which looks almost spherical. The height differences between the uppermost point of the cells and the coalescence point of the junctions further increased to (4.3 ± 0.4) µm. The morphological

changes of the cells as described above are strong and characteristic indicators for the onset of cell necrosis due to the unphysiological pH values.45,46 SECM Mediator Selection. A suitable redox mediator for SECM studies with living cells has to be nontoxic and must exhibit a reliable electrochemical behavior at Pt UMEs. Besides the frequently used SECM mediators [Ru(NH3)6]Cl3,47 FcMeOH, FcCOOH, and K4[Fe(CN6)], some other cyanometallates such as K4[W(CN8)] · 2H2O and K4[Os(CN6)] · 3H2O were evaluated.48 NRK cell monolayers were exposed to the mediator solutions (1 mM mediator in PBS++ containing 1 g/L Dglucose) at room temperature. After 6 h of incubation, a Calcein-AM based live/dead assay was performed to determine the cytotoxicity of either mediator. Among the mediators being tested, K4[Fe(CN6)] and K4[Os(CN6)] · 3H2O were found to be toxic to NRK cells. Since redox mediators should give a stable electrochemical signal at least over the time of the image acquisition, chronoamperometric measurements of the remaining mediators were performed. In addition, the dependency of SECM imaging on the mediator properties is an important aspect. To avoid misinterpretations of the SECM images, the cell monolayers were investigated using two very different redox mediators, [Ru(NH3)6]Cl3 and FcMeOH. Both mediator systems have different redox properties, and FcMeOH exhibits an amphiphilic character that might lead to an accumulation in cell membranes. A mathematical comparison of SECM images recorded with both mediator systems showed only negligible differences. In conclusion, both mediator systems are suitable to study topographical changes of NRK cell monolayers. FcMeOH shows a good long-term stability of the chronoamperometric signals in PBS++ and was selected for further investigations in the context of the present study. Imaging of NRK Cell Monolayers by SECM. Prior to image acquisition, the electrochemical cell with the fixed glass coverslip was leveled by recording probe approach curves at three positions down to currents of 75% of the steady-state current I0. The UME was positioned over the cell monolayer, and approach curves were carefully recorded down to currents of 60% of I0. Figure 3 illustrates a family of approach curves for different positions over the monolayer. It was found that the shape of the approach curve was dependent on the specific position of the UME over

Figure 4. Highly resolved SECM image (A, false colors; B, 3D) of a confluently grown NRK cell monolayer. The image was recorded in 1.5 mM FcMeOH using a 300 nm radius UME with a tip velocity of 10 µm/s. 172

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

Figure 5. SECM images (A, C, E, false colors; B, D, F, 3D) of an NRK cell monolayer during a localized increase in pH. The morphology of the monolayer before the chemical attack of a single living target cell is illustrated in A/B, whereas C/D and E/F were recorded 10 and 20 min after the electrochemically induced stimulus. Other experimental parameters are as in Figure 4.

the substrate. Occasionally the approach curves were less steep than expected. This deviation from theory has been described elsewhere for the case of PC12 cells47 and was explained by differences in cellular curvature. Assuming a hemispherical cell topography approach, curves will show deviations of the negative feedback current from theory which holds only for perfectly flat substrates. The UME was stopped at currents of 60% of I0 to avoid mechanical interactions with the cells. Line scans were performed to characterize an appropriate section within the monolayer. The acquisition of SECM images was started when these scans gave reproducible results and showed no parallel offset in scan direction during repeated scans due to movement of the substrate.

Figure 4 presents a typical SECM image of a confluently grown NRK cell monolayer using a 300 nm radius UME. Investigating a detail of 40 × 40 µm2 and applying a tip velocity of 10 µm/s, it took about 10 min to generate one micrograph. Studying cell monolayers in the constant-height mode is advantageous over the alternative constant-distance methods with integrated distance control. A tip crash can easily occur particularly in the context of high-resolution imaging of cell-cell contacts in the constant-distance mode, as these topographic details are located about 0.3-1.0 µm beneath the uppermost point of the cell bodies. Since the lateral distances of the cell-cell contacts are about 1-3 µm, it is difficult to scan the cell-cell contacts with the UME in a sub-micrometer distance without an interaction of the glass insulation with the cell bodies. As Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

173

illustrated in Figure 4, single cells can be clearly identified and imaged by constant-height scanning in the form of false color (A) and 3D plots (B). The typical cobblestone morphology of NRK cells is perfectly illustrated. It can be seen that there are slight inhomogeneities within the cell monolayer. The four yellow colored cells shown in the bottom and middle part of Figure 4 (A) lie in one plane, whereas the top left cell lies in a slightly deeper plane which is indicated by the blue color. It is important to note that even cell-cell contacts between two adjacent cells, which are located at different z levels, could reliably be imaged. The resolution and quality of the SECM image favorably compares with phase contrast (Figure 1) and confocal laser scanning micrographs (Figure 2). Chemical Attack of One Single Living NRK Cell within the Monolayer. Besides high-resolution imaging, a sub-micrometer UME can also be used to microstructure a substrate. We present the chemical attack of one single living cell integrated within a cell monolayer. This approach enables the treatment of a single target cell relative to reference cells in the surrounding area which serve as an internal standard for the cell manipulation. The attack of the target cell was performed by local generation of hydroxide ions at the UME (see Supporting Information for the chronoamperogram). There are, however, challenges regarding the reliability of the UME response, since the image acquisition and the chemical attack have to be done with the same UME. Figure 5A/B shows a micrograph that reflects the original state of a confluent NRK cell monolayer under normal physiological conditions. In this case, all cells are on an equal height level as indicated by similar colors in the central parts of the cells. After the image acquisition, the UME was positioned centrally above the target cell, which is located in the middle of Figure 5A/B with the coordinates (14 µm, 14 µm). The chemical attack was carried out by applying a potential of -1.1 V for 160 s to generate hydroxide ions. Since the amount of generated hydroxide ions was very small (about 135 fmol were calculated from the chronoamperogram), a PBS buffer with reduced buffer capacity was used. The buffer concentration was lowered to 1 mM. In this way a very localized increase in pH just in the vicinity of a single cell could be realized. The smooth current trace indicates that no hydrogen gas bubbles were evolved at the UME. This is attributed to the very fast mass transport at the UME. The same region of the monolayer was subsequently imaged again using the UME. Care had to be taken to adjust the z coordinate of the UME to the changed height of the target cell by recording a new probe approach curve down to currents of 60% of I0. The electrode characteristics were altered due to the application of the highly negative potential used for hydroxide (43) Olsson, O. Z. Anorg. Allg. Chem. 1914, 88, 49–53. (44) Krauss, F.; Schrader, G. J. Prakt. Chem. 1928, 119, 279–286. (45) Berghe, T. V.; Vanlangenakker, N.; Parthoens, E.; Deckers, W.; Devos, M.; Festjens, N.; Guerin, C. J.; Brunk, U. T.; Declercq, W.; Vandenabeele, P. Cell Death Differ. 2010, 17, 922–930. (46) Alberts, B.; Johnson, A.; Lewis, J.; Raff, M.; Roberts, K.; Walter, P. Molecular Biology of the Cell, 2nd ed.; Garland Science: New York, 2002. (47) Liebetrau, J. M.; Miller, H. M.; Baur, J. E. Anal. Chem. 2003, 75, 563– 571. (48) Matysik, F.-M.; Khoshtariya, D. E.; Billing, R. Port. Electrochim. Acta 1997, 15, 151–158. (49) DeMello, W. C.; Janse, M. J. Heart Cell Communication in Health and Disease; Kluwer Academic Publishers: Dordrecht, 1997.

174

Analytical Chemistry, Vol. 83, No. 1, January 1, 2011

ions generation. This is indicated by the fact that the steadystate current for the oxidation of FcMeOH was lower than in initial recording. Despite the aging of the electrode, the morphological response of the treated cell is visualized in the SECM image. Figure 5C/D was recorded 10 min after the pH increase. It can be seen that all relative positions of the peripheral cells in the xy plane remained unchanged. The shape and particularly the height of the target cell have changed, whereas no height difference among the reference cells could be detected. Obviously only the target cell was selectively altered. The target cell began to change its former rectangular shape to a more semispherical one. This cell response is typical for a necrotic cell which tends to swell due to an increased water influx.45,46 In addition, the junctions of the necrotic cell to adjacent cells were affected. For further investigations of the target cell, the UME had to be moved to a higher position in order to avoid a tip crash with the cell. Figure 5E/F was recorded 20 min after the pH increase. The shape difference between the target cell and the control cells is even more evident. In comparison to the previous image, the necrotic target cell has gained considerable volume and shows a cupola-like morphology, as expected due to an imminent separation out of the monolayer. Despite the clear necrotic response of the target cell, all neighboring cells remain essentially unaffected even though they are in direct contact. This image illustrates nicely the well-known proverb that cells live together and die alone.49 CONCLUSIONS The present study demonstrates that SECM performed in the constant-height mode is an ideally suited technique to image individual cells within a confluent cell monolayer. The use of UMEs with effective radii between 200 and 600 nm enabled an imaging performance of SECM characterized by subcellular lateral resolution which compares favorably with images recorded by phase contrast or confocal laser scanning microscopy. Moreover, as an additional and unique experimental option of SECM, the same UME used for imaging could be used to address and attack individual cells within the cell monolayer by a locally confined electrogeneration of hydroxide ions. In this way it was possible to trigger necrosis in an individual cell and to perform a subsequent imaging of the corresponding morphological changes of the target cell. In a broader context this experimental approach is very promising for more sophisticated single cell studies because neighboring cells within the cell monolayer can serve as reference objects to enhance the reliability of analytical interpretations. ACKNOWLEDGMENT S.B. acknowledges gratefully the Fonds der Chemischen Industrie for providing a Kekule´ grant. SUPPORTING INFORMATION AVAILABLE Additional information as noted in the text. This material is available free of charge via the Internet at http://pubs.acs.org.

Received for review August 13, 2010. Accepted November 14, 2010. AC1021375