Stress relaxation microscopy: Imaging local stress in cells

ARTICLE IN PRESS Journal of Biomechanics 43 (2010) 349–354 Contents lists available at ScienceDirect Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com Stress relaxation microscopy: Imaging local stress in cells Susana Moreno-Flores a, Rafael Benitez b, Maria dM Vivanco c, Jose Luis Toca-Herrera a, a b c Biosurfaces unit, CIC BiomaGUNE, Paseo Miramón 182, 20009 San Sebastia n-Donostia, Spain Department of Mathematics, Centro Universitario de Plasencia, Universidad de Extremadura, Avda. Virgen del Puerto 2, 10600 Plasencia, Spain Cell Biology & Stem Cells Unit, CIC BioGUNE, Parque tecnológico de Bizkaia, Ed. 801A, 48160 Derio, Spain a r t i c l e in fo abstract Article history: Accepted 21 July 2009 Biomechanics is gaining relevance as complementary discipline to structural and cellular biology. The response of cells to mechanical stimuli determines cell type and function, while the spatial distribution of mechanical forces within the cells is crucial to understand cell activity. The experimental methodologies to approach cell mechanics are diverse but either they are effective in few cases or they rule out the innate cell complexity. In this regard, we have developed a simple scanning probebased methodology that overcomes the limitations of the available methods. Stress relaxation, the decay of the force exerted by the cell surface at constant deformation, has been used to extract relaxational responses at each cellular sublocalisation and generate maps. Surprisingly, decay curves exerted by test cells are fully described by a generalized viscoelastic model that accounts for more than one simultaneously occurring relaxations. Within the range of applied forces (0.5–4 nN) a slow and a fast relaxation with characteristic times of 0.1 and 1 s have been detected and assigned to rearrangements of cell membrane and cytoskeleton, respectively. Relaxation time mapping of entire cells is thus promising to simultaneously detect non-uniformities in membrane and cytoskeleton and as identifying tool for cell type and disease. & 2009 Elsevier Ltd. All rights reserved. Keywords: Cells Atomic force microscopy Mechanical properties Stress relaxation 1. Introduction Cell biomechanics is becoming a diagnostic tool in cell biology. Cell response to mechanical stimuli are distinctive of cell type (Elson, 1998), cell function (Huang et al., 2004) and cell damage (Rotsch and Radmacher, 2000) and are believed to be strongly dependent on the cytoskeleton (Elson, 1998). Structural proteins such as keratins especially confer mechanical resistance on epithelial cells (Coulombe and Omary, 2002); human diseases like most of the haemolytic anemias and cirrhosis are associated with elasticity loss in red blood cells and hepatocytes, respectively. Additionally, carcinoma cells exhibit anomalous compressibility or elasticity when compared to healthy cells of the same type (Wu et al., 2000; Darling et al., 2007; Li et al., 2008). Few are the ways to impose a mechanical stimulus to a cell and observe its response: In the so-called transient experiments sudden stress induces creep in cells that can be monitored using optical microscopy. Based on that, techniques such as micropipette aspiration (Hochmuth, 2000) and microplate manipulation (Thoumine and Ott, 1997) are particularly relevant in the study of whole-cell creep mechanics of non-adherent, light-adherent or suspended cells. Additionally, dynamic shear stress induces  Corresponding author. Tel.: + 34 943 00 53 13; fax: + 34 943 00 53 01. E-mail address: jltocaherrera@cicbiomagune.es (J.L. Toca-Herrera). 0021-9290/$ - see front matter & 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2009.07.037 time-dependent deformations that can be detected in microrheology (i.e. dynamic) experiments. In this case, laser-track (Yamada et al., 2000) and magnetic probe-based (Valberg and Albertini, 1985; Bausch et al., 1998) techniques allow studying the viscoelasticity of the intracellular space and the cell surface by monitoring the displacement of internalized and surface-attached microparticles, respectively. Of all the different techniques currently available to test cell response to mechanical stimuli, the cell poker (Petersen et al., 1982) and scanning probe (SP) based-force spectroscopy (Weisenhorn et al., 1993) are most suitable to study the local behaviour on adherent cell surfaces and tissues (Huang et al., 2004). Deformation and stress are applied normal to the cell surface either with a microsized glass stylus or a submicro-sized silicon (nitride) tip positioned at a certain location. Due to the smaller tip size, the SP-based technique provides better spatial resolution and has been mainly applied to the study of the elastic stress–strain behaviour of cells and the obtention of elastic moduli (Butt et al., 2005). However, modelling has been greatly restricted in these cases. Except for SP-based microrheological studies (Alcaraz et al., 2003), cells have been conceived as purely elastic, homogeneous materials that are subjected to small (10– 100 nm), sudden deformations. Typical probing areas are submicrometer-sized, which questions the validity of cells being homogeneous bodies normal to and along the cell surface. The scanning probe can sense various cell components, especially at ARTICLE IN PRESS 350 S. Moreno-Flores et al. / Journal of Biomechanics 43 (2010) 349–354 high deformations. All these components (cell membrane, actin cortex, other cytoskeletal components) may respond differently to probe-induced stimuli. Additionally, probing the cell at different locations provides information on the lateral distribution of mechanical responses, which may in turn differ (Radmacher et al., 1996). In this work we describe an SP-based imaging methodology, denoted as stress relaxation microscopy (STREM) that generates stress relaxation maps of complete cells, which have been subjected to larger range of deformations than those reported to date. Our data analysis, in contrast to previously reported stress relaxation assays (Darling et al., 2007; Okajima et al., 2007) takes into account both cell three-dimensional heterogeneity and cell viscoelasticity. To test our method, we have used human breast cancer cells (MCF-7). This epithelial breast adenocarcinoma cell line is one of the most frequently used model systems to study breast cancer. The MCF-7 cell line was derived from a pleural effusion of a patient with metastatic breast cancer (Levenson and Jordan, 1997), however, this line normally does not metastasize. The relaxation of the force exerted by MCF-7 cells on SPcantilevers after the application of a sudden deformation has been characterized within a wide range of applied deformations and cell locations. A multicomponent viscoelastic model has been used to interpret the data, to extract mechanical properties and to correlate the latter to cell topology. 2. Materials and methods 2.1. Sample preparation MCF-7 cells were grown at 37 1C and 5% CO2 in Dulbecco’s modified eagle medium (DMEM, Sigma) supplemented with 8% fetal bovine serum (FBS, Sigma), 2% 200 mM L-glutamine, 0.4% penicilline/streptomicine (PEN/STREP, Sigma). For force measurements, the cells were subcultured on borosilicate glass coverslips (diameter 24 mm and 0.16 mm thickness) at a density of 25, 15 and 10 K/ml and left to incubate for 1, 2 and 3 days, respectively. Prior to force measurements, the cells were washed in CO2-independent cell medium (Leibowitz medium, L15, Sigma) and measured in the same medium at 37 1C. 2.2. Force–time curves Measurements were carried out on different cell clusters for the same sample with a Nanowizard II (JPK Instruments, Germany) coupled with a transmission optical microscope (Axio Observer D1 Zeiss, Germany). Gold backside coated SiN cantilevers of nominal spring constant of 0.006 N/m (MLCT, Veeco Instr., USA) were used for both cell imaging and force measurements. The cantilevers were previously cleaned in acetone and ethanol to remove impurities and their actual spring constants that ranged from 0.01 to 0.019 N/m, were evaluated in air by the thermal method. The cell-coated glass substrates were then mounted in a lowvolume cell incubator (Biocell, JPK Instruments, Germany), with 400 ml of L15 cell medium and thermalized at 37 1C. Individual force–time curves were recorded on each of the 2–8 cell clusters at a speed of 5 mm/s and at maximum loads of 0.5, 1, 2, 3 and 4 nN and at least two different cell positions. Application of loads higher 4 nN did not guarantee cell resistance in all cases. Force relaxation was registered at constant height mode, where the contact time was set to 2 s. Together with the force relaxation, approach and withdrawal curves were also obtained. For STREM, the evaluated area was divided either in 25  25 or 30  30 pixels and in each pixel a force–time curve was registered. The maximum load for mapping was kept constant to 2 nN, the rate and the contact time set to 5 mm/s and 2 s, respectively. To obtain a STREM map takes approximately 90 min under these conditions. In all cases, optical micrographs of the cells were taken before and after the experiment to rule out the possible tip-induced cell damage. 2.3. Force curve analysis Individual force–time curves were analysed using an own-deviced Matlabs routine. First of all the force–time curves are tilt-corrected to compensate baseline drift. The fitting range is automatically selected from the time derivative of the piezo displacement, which is approximately zero in the force-relaxation range (constant piezo position with time). The corresponding force curve is fitted to a double exponential using the Levenberg–Marquart algorithm. Deformations are calculated from the difference between the piezodrive positions of the contact point and the maximum load of the force curve on the cell. The difference is then corrected by subtracting the displacement of the cantilever against the hard substrate for the same maximum load. The contact point is identified from corrected force curves in the force–displacement representation as the smallest displacement where the derivative of the force is no longer negative (baseline points fluctuate around zero force with positive and negative derivatives). The identification is in most cases easy, except for a few cases where the tilt-correction of the baseline was not appropriately done and manual identification was required. Manual identification of the contact point as the baseline end presented no problems. STREM maps are also obtained from an own deviced Matlabs routine that implements the individual force curve analysis. The analysis results of each curve (normal tension decay, the cell deformation and the relaxation time during contact time) are stored together with the offset position (the space position at which it was performed) and plotted as coloured scale pixel. 3. Results 3.1. Force–time curves—compressive force relaxation curves We have registered as a function of time the force exerted by MCF-7 cells (Fig. 1a) on the SP cantilever during tip approach, tipcell contact, and tip withdrawal (Fig. 1b). At t0 starts the contact time, which is the time when the tip remains in contact with the cell, and the force steadily decreases with time from an initial value (the maximum load, see Fig. 1b). The force eventually reaches a plateau if contact is maintained sufficiently long (curve 1 in Fig. 1b). This observed force decay is symptomatic of probing non-elastic bodies, in opposition to purely elastic ones like the glass substrate, which does not exhibit such behaviour (curve 2 in Fig. 1b). We have additionally observed that the force–time dependence in MCF-7 cells – under the studied loads – obeys a double-exponential decay (dashed line in Fig. 1b). An exponential force decay is the typical relaxation response of certain linear, isotropic viscoelastic bodies subjected to a constant deformation (Riande et al., 2000). Soft biological tissues such as kidney tissue or costal cartilage exhibit such behaviour (Mattice et al., 2003; Jin and Lewis, 2004). According to the spring-dashpot model developed by Maxwell, which consists of an elastic spring connected in series with a viscous dashpot, a material characterized by a compressive elastic modulus E, and a viscosity Z exhibits a force response to a sudden and constant deformation set at time t0 that can be defined as follows: F ¼ A exp t¼ Z E   ðt  t0 Þ t ð1Þ where A is the force amplitude of the relaxation and t the relaxation time. This relation holds as long as both the deformation and the contact area (region of the material along which the mechanical deformation is applied) are constant.1 Accordingly, a generalized Maxwell model consisting of N parallely arranged Maxwell elements describes multiexponential decays in 1 Maxwell model predicts that the compressive modulus G depends with time according to: G(t)= G exp(  ((t  t0)/t)), where G= s/e, being s the compressive stress (force divided by contact area) and e the relative deformation. If both deformation and contact area are constant, the time relation for the compressive modulus also holds for the compressive force with the same relaxation time, t. Though we do not have experimental evidence to confirm that this is the case in our experiments, we can estimate the maximum change of deformation along the contact time to be 7–12.9% and 3–6.7% of the total cytoplasmic and nuclear instantaneous deformations, respectively. These values are calculated from changes in the cantilever deflection (as obtained from the calibration curves against a hard substrate) that would correspond to the force decays obtained at constant height. As a first approximation we assume these changes to be small enough to question the validity of linear viscoelastic theory. ARTICLE IN PRESS S. Moreno-Flores et al. / Journal of Biomechanics 43 (2010) 349–354 2 t0 force (nN) 1 351 2 1 25µm time (s) 10µm 25µm 800 pN 25µm Fig. 1. (a) Differential interference contrast (DIC) optical micrograph of a cluster of four MCF-7 cells where nuclear regions are distinguishable; (b) force–time curve registered during tip approach (times smaller than t0), tip-cell contact (t0–4 s) and tip withdrawal (times longer than 4 s). Curves are performed on the nuclear region of one of the MCF-7 cells (location depicted as 1 in Fig. 1a, curve 1) and on the glass substrate (location depicted as 2 in Fig. 1a, curve 2); (c) height SP micrograph of the same cluster indicates that the maximum cell heights are attained on the nuclear regions (Z 5 mm) and (d) SP cantilever deflection micrograph showing the details of cytoskeletal fibers at cell edges. where A0 accounts for the instantaneous (purely elastic) response. Each element thus contributes to the overall response as an individual relaxational process. In our experiments the cantilever position was kept constant during the contact time and the forces and deformations are applied normal to the cell surface. Contact area is assumed to be constant along with deformation. Therefore the force change we register on cells during this time is thus interpreted as force relaxation under constant compression with N=2 simultaneously occurring processes. A1 and A2 are the corresponding decay amplitudes and t1 and t2 their relaxation times. On the glass substrate, no noticeable force relaxation is detected and thus is characterized by zero decay amplitudes (A1 = A2 =0). amplitude of the force decay and the relaxation times are plotted against the cell deformation (Fig. 2b) and the initial load (Fig. 2c), respectively. The behaviour of the force decay-to-cell deformation differs on the nuclear and cytoplasmic (i.e., perinuclear) regions of the cell. Fig. 2b shows a linear dependence on both regions within the range of maximum loads evaluated; however, the nuclear region is prone to larger deformations than the perinuclear region, and therefore the data of the former extends to higher deformation values. Hence, force decay-to-cell deformation ratio (i.e., the normal tension decay) is larger on the perinuclear region than on the nuclear region, which denotes that this magnitude depends on the subcellular localisation. Fig. 2c, shows that the fast and slow relaxation times, t1 and t2, respectively, differ one order of magnitude, t1 being in the range of 100 ms and t2 in the range of seconds. Surprisingly their values neither depend substantially on the maximum load applied nor on the subcellular localisation. However, evaluating at random positions does not provide a complete picture of the cell response, as we will show in the next section. 3.2. Dependence of the relaxation on maximum loads and subcellular localisation 3.3. Mapping compressive force decays and relaxation times: STREM images When performing the experiments at different loads from 0.5 to 4 nN, the time dependence of the force during the contact time always decays in a double-exponential fashion. Fig. 2a shows as black traces the experimental curves performed on the nuclear region of an MCF-7 cell and the corresponding double-exponential fitting (red curves) according to Eq. (2). The quality of the fits is good as can be seen in Fig. 2a, which allowed obtaining both the overall amplitude of the force decay (A1 +A2) and the relaxation times (t1 and t2) in each case. Analogous experiments on the cytoplasmic region of the cells allowed comparison of relaxational processes set at different subcellular localisations. The result of this comparison is shown in Fig. 2b and c, where the Magnitudes that can be extracted on each cellular location, such as normal tension decays, t1 and t2, can be mapped to obtain a three-dimensional view of the relaxational processes on the complete cell surface. In this case, the maximum load applied is unique for all positions, and force–time curves are taken while the SP cantilever is step-scanning an individual cell or a cell cluster. As an example, Fig. 3 shows STREM maps of two individual MCF-7 cells. The maximum load has been set equal to 2 nN, high enough to clearly distinguish both relaxations and low enough to be within the linear regime (see Fig. 2b) and ensure cell integrity. Distribution of cell heights accounts for image contrast in Fig. 3a, where nuclear regions appear thick and bulgy while perinuclear heterogeneous materials as follows:   N X ðt  t0 Þ Ai exp FðtÞ ¼ A0 þ i¼1 ti ¼ ti Zi Ei ð2Þ ARTICLE IN PRESS 352 S. Moreno-Flores et al. / Journal of Biomechanics 43 (2010) 349–354 Fig. 2. (a) Force–time curves (black traces) on position 1 (see Fig. 1) at different maximum loads (0.5, 1, 2, 3 and 4 nN), the red curves being the double-exponential fittings. Inset: viscoelastic model that consists of a spring connected in parallel with two Maxwell elements; (b) total force decay versus local deformation for two nuclear and two cytoplasmic (perinuclear) regions of the same cell, the straight lines are fits to the experimental data and (c) relaxation times, t1 and t2, for those regions. Fig. 3. Height (a) and STREM (b–d) images of individual MCF7 cells with the corresponding histograms. The height map is a distribution of heights at constant force (1 nN), cell heights extend from 1 mm (perinuclear regions) to 5 mm (nuclear regions); (b) map and histogram of the ratio between force decay and deformation, the normal tension decay; (c) map and histogram of the fast relaxation time; (d) map and histogram of the slow relaxation time. Histogram binning (i.e. width of columns) is comparable to the parameter error. Black pixels outside the cells refer to substrate, where no relaxation or deformation occurs (no numbers associated). regions appear flatter. Height SP micrographs (Fig. 1c) on MCF-7 clusters confirm this fact. The force decay amplitudes in Fig. 3b have been divided by the cell deformation at each pixel to obtain normal tension decays and to allow comparison between the different cell localisations. Image contrast in Fig. 3b is thus indicative of biomechanical disparity and it correlates with cell morphology: the bulgy, nuclear regions of the cells exhibit larger deformations and thus appear darker than the perinucleus. Figs. 3c and d show maps of the fast (t1) and slow (t2) relax- ation times, respectively. These measurements illustrate the importance of mapping relaxation times of the whole cell in opposition to the results shown in Fig. 2c, where measurements were performed on a few cell locations. Fig. 2c would lead to the erroneous conclusion that the cell time response is positionally homogeneous. On the contrary Figs. 3c and d show the complexity of the cell response, which is positionally non-homogeneous and more comprehensively characterized by mapping and histograms. We thus expect that a statistical description of the relaxational ARTICLE IN PRESS S. Moreno-Flores et al. / Journal of Biomechanics 43 (2010) 349–354 353 behaviour in cells will be distinctive of the cell type and function. 4. Discussion 4.1. Overall force decay versus deformation Overall force decay depends linearly on cell deformation in both nuclear and perinuclear regions (Fig. 2b). However, the perinuclear region is usually less compliant to deformation than the nuclear region (Weisenhorn et al., 1993; Radmacher et al., 1996), that is also observed in our experiments. Therefore the normal tension decay (A1 + A2)/Dl (Waugh and Evans, 1976), the slope of the overall force decay versus deformation, depends on subcellular localisation. On perinuclear regions, the slopes are approximately twice as much as those obtained on nuclear regions2 (Fig. 2b). This numerical discrepancy may be strongly linked to the relative higher compressibility of the nuclear region in comparison to the perinuclear. The former has a higher content of intracellular fluid and the volume density of cytoskeletal fibers is lower than on the perinucleus, which accounts for its higher compressibility. Mapping normal tension decays on a complete cell is thus most illustrative of cell morphology and response distributions (Fig. 3b). 4.2. Bi-exponential decay: possible reasons The observed biexponential decay may have two possible reasons. Upon cell compression, compressive and shear forces may occur, which result in two relaxational processes if decoupled. Alternatively, it is reasonable to expect that two relaxations may also account for probing two different cell environments. Let us first consider the first case. Our probe has a pyramidal shape. On contacting the cell, the local cell shape may be deformed around the pyramidal tip. In this case, indentation would take place and both compressive and shear deformations occur upon the contact area that extends beyond the apex of the tip (compressive) to the pyramidal sides (shear). Hence both compressive and shear forces would contribute to the detected relaxation. The two relaxational processes found may thus account for the existence of these two types of forces:   ðt  t0 Þ FðtÞ  A0 ¼ Fcompressive ðtÞ þ Fshear ðtÞ ¼ Acompressive exp tcompressive   ðt  t0 Þ ð3Þ þ Ashear exp tshear each one characterized by a decay amplitude and a relaxation time. In this case, the contribution (i.e., amplitudes) of these two terms to the overall force decay should certainly depend on probe geometry. A pyramidal-nanosized tip indents more than a microsized spherical colloid and therefore should induce a higher shear force. The contribution of the second term in Eq. (3) to the overall force decay, A*shear =Ashear/(Atcompressive +Ashear), should thus be greater in the case of a pyramidal probe than in the case of a colloidal probe. In other words, A*shear should be larger for tips than for colloids and (A*shear)(tip)/(A*shear)(colloid) larger than one, irrespectively of the applied load. In our experiments we obtain two amplitudes, A*1 =A1/(A1 + A2) and A*2 =A2/(A1 + A2), which cannot a priori be attributed to shear or compressive force decays. We 2 The slopes on the nuclear region are 1.207 0.07 and 0.93 7 0.06 mN/m and on the perinuclear region the slopes amount to 2.77 0.1 and 2.017 0.2 mN/m. Fig. 4. Force decay amplitude ratios for different probe geometries. Tip: silicon nitride square pyramids with 101 7 11 nm apex radius (measured by scanning electron microscopy, SEM). Colloid: silicon bead of 8 mm diameter (also measured by SEM). The magnitude A*i (i= 1,2) refers to the ratio Ai/(A1 + A2). Within the experimental error both amplitude ratios are indistinguishable from one another. The experiments were performed on the nuclear area of MCF-7 cells. Results show statistical averages from 13 (pyramidal tip) and 10 (colloidal probe) cells. thus performed stress relaxation experiments with pyramidal and colloidal tips on nuclei of MCF-7 cells and plotted (A*1)(tip)/ (A*1)(colloid) and (A*2)(tip)/(A*2)(colloid) as a function of the applied load (Fig. 4). The results show that, within the experimental error, the calculated ratios are close to one, meaning no difference in the contributions to the overall force decay. The biexponential decay is therefore not due to the existence of decoupled shear and compressive relaxational processes. In the second case, when shear relaxation does not occur or cannot be decoupled from compressive relaxation, the two observed decays may account for relaxational processes in two different cell environments and thus the equation has the following form:   ðt  t0 Þ FðtÞ  A0 ¼ Fcc1 ðtÞ þFcc2 ðtÞ ¼ Acc1 exp tcc1 þ Acc2 exp   ðt  t0 Þ tcc2 ð4Þ where cc1 and cc2 stand for cell component 1 and 2, respectively. The time scale of the fast relaxation process in our experiments, t1, lies between 0.1 and 0.3 s, which greatly resembles the response times of red cell membranes (Waugh and Evans, 1976) and of macrophages (t = 0.218 s) (Bausch et al., 1999). The slow relaxation, t2, is of the order of seconds, which is within the time scale of cytoskeletal rearrangements (Theret et al., 1998; Yamada et al., 2000). Assigning the fast and slow relaxations to membrane and cytoskeletal responses may be ad hoc; however it agrees with the fact that both relaxations proceed with similar time scales on both nuclear and perinuclear regions of the cell (Fig. 2c). The two outermost components that completely surround the cell are the cell membrane and the cytoskeletal cortex (Alberts et al., 1994; Anathakrishnan et al., 2006) that are most likely sensed by the scanning probe. STREM images show that it is possible to exploit the capabilities of the scanning probe technique to address local mechanical processes within the cell and obtain relaxation maps. The values of t1 and t2 are similar along their surface, though not identical. Nonuniformities in the relaxation time maps appear randomly distributed in the case of the t1 map, which may reflect the ARTICLE IN PRESS 354 S. Moreno-Flores et al. / Journal of Biomechanics 43 (2010) 349–354 structural heterogeneity of the cell surface. In this regard, the presented methodology could provide, at higher lateral resolution local structural ‘‘anomalies’’ on the cell surface (e.g., protein clusters, gap junctions or local adhesion points) (Plattner and Hentschel, 2006). In the t2 map a double distribution of relaxation times was easily observed, the higher values are mainly found in the regions close to the cell edges, where the volume density of fibers are high (see Fig. 1d). STREM can thus provide access to regions within the cell characterized by distinct time responses. 5. Conclusions We have developed an alternative, SP-based methodology, which provides access to maps of local mechanical parameters and allows characterizing complete cells. We have used the MCF-7 cell line as a model system to which we have applied local deformations with an SP cantilever. The force decay has been fully characterized as a function of time. MCF7 cells relax according to two simultaneously occurring processes that may involve cell membrane and cytoskeletal rearrangements. Our work illustrates the importance of STREM mapping in the biomechanical characterization of cells. Thorough mapping of normal tension decays and time relaxations of complete cells can provide a comprehensive view of the complexity of cell relaxational biomechanics and eases to link the local mechanical behaviour with cell morphology. As a matter of fact, STREM could be a potential tool to localise gap-junctions and caveolae in invitro cells. The authors envisage STREM as a developing technique with wide applicability though it requires refinement. The influence of the elasticity of the scanning probe on the observed relaxations is still an open question. Diagnosing cell activity and dysfunction are among the biomedical applications of STREM, although its utility is not restricted to cells. (Bio)polymer films and scaffolds, liposomes and tissue among others are also viscoelastic materials that can be fully characterized by this method. Conflict of interest statement This letter is to confirm that the authors of the manuscript entitled Stress Relaxation Microscopy: Imaging Local Stress In Cells does not have any conflict of interest. Acknowledgments The authors thank Verónica Saravia for technical assistance in the experiments with colloidal probes and also to Marco Piva for help with cell culture. We additionally thank Dr. Francisco J. del Castillo, Dr. Kathrin Melzak and Prof. Helmuth Möhwald for reading and useful comments. SMF, MdMV and JLTH thank the ETORTEK programme of the Basque Government for financial support. References Alberts, B., Bray, D., Lewis, J., Raff, M., Roberts, K., Watson, J.D., 1994. Molecular Biology of the Cell. Garland Science, New York, p. 135. Alcaraz, J., Buscemi, L., Grabulosa, M., Trepat, X., Fabry, B., Farre, R., Navajas, D., 2003. Microrheology of human lung epithelial cells measured by atomic force microscopy. Biophysical Journal 84, 2071–2079. Anathakrishnan, R., Guck, J., Käs, J., 2006. Cell mechanics: recent advances with a theoretical perspective. In: Pandalai, S.G. (Ed.), Recent Research Developments in Biophysics, Vol. 5. Transworld Research Network, Kerala, pp. 39–69. Bausch, A.R., Zieman, F., Boulbitch, A.A., Jacobson, K., Sackmann, E., 1998. Local measurements of viscoelastic parameters of adherent cell surfaces by magnetic bead microrheometry. Biophysical Journal 75, 2038–2049. Bausch, A.R., Moller, W., Sackman, E., 1999. Measurement of local viscoelasticity in living cells by magnetic tweezers. Biophysical Journal 76, 573–579. Butt, H.J., Cappella, B., Kappl, M., 2005. Force measurements with the atomic force microscope: technique, interpretation and applications. Surface Science Reports 59, 1–152. Coulombe, P.A., Omary, M.B., 2002. ‘‘Hard’’ and ‘‘soft’’ principles defining the structure, function and regulation of keratin intermediate filaments. Current Opinion in Cell Biology 14, 110–122. Darling, E.M., Zauscher, S., Block, J.A., Guilak, F., 2007. A thin-layer model for viscoelastic, stress-relaxation testing of cells using atomic force microscopy: do cell properties reflect metastatic potential?. Biophysical Journal 92, 1784–1791. Elson, E.L., 1998. Cellular mechanics as an indicator of cytoskeletal structure and function. Annual Review of Biophysics and Biophysical Chemistry 17, 397–430. Hochmuth, R.M., 2000. Micropipette aspiration of living cells. Journal of Biomechanics 33, 15–22. Huang, H., Kamm, R.D., Lee, R.T., 2004. Cell mechanics and mechanotransduction: pathways, probes and physiology. American Journal of Physiology 287, 1–11. Jin, H., Lewis, J.L., 2004. Determination of Poisson’s ratio of articular cartilage by indentation using different-sized indenters. Journal of Biomechanical Engineering 126, 138–145. Levenson, A.S., Jordan, C.V., 1997. MCF-7: the first hormone-responsive breast cancer cell line. Cancer Research 57, 3071–3078. Li, Q.S., Lee, G.Y., Ong, C.N., Lim, C.T., 2008. AFM indentation study of breast cancer cells. Biochemical and Biophysical Research Communications 374, 609–613. Mattice, J.M., Lau, A.G., Oyen, M.L., Kent, R.W., 2003. Spherical indentation loadrelaxation of soft biological tissues. Journal of MaterialMaterials Research 21, 2003–2010. Okajima, T., Tanaka, M., Tsukiyama, S., Kadowaki, T., Yamamoto, S., Shimomura, M., Tokumoto, H., 2007. Stress relaxation of HepG2 cells measured by atomic force microscopy. Nanotechnology 18, 084010,1–0840105. Petersen, N.O., McConnaughey, W.B., Elson, E.L., 1982. Dependence of locally measured cellular deformability on position on the cell, temperature, and cytochalasin B. Proceedings of the National Academy of Sciences USA 79, 5327–5331. Plattner, H., Hentschel, J., 2006. Taschenlehrbuch Zellbiologie. Georg Thieme Verlag, Stuttgart, p. 515. Radmacher, M., Fritz, M., Kacher, C.M., Cleveland, J.P., Hansma, P.K., 1996. Measuring the viscoelastic properties of human platelets with the atomic force microscope. Biophysical Journal 70, 556–567. Riande, E., Dı́az-Calleja, R., Prolongo, M.G., Masegosa, R.M., Salom, C., 2000. Polymer Viscoelasticity. Stress and Strain in Practice. Marcel Decker, New York, p. 879. Rotsch, C., Radmacher, M., 2000. Drug-induced changes of cytoskeletal structure and mechanics in fibroblasts: an atomic force microscopy study. Biophysical Journal 78, 520–535. Theret, D.P., Levesque, M.J., Sato, M., Nerem, R.M., Wheeler, L.T., 1998. The application of a homogeneous half-space model in the analysis of endothelial cell micropipette measurements. Journal of Biomechanical Engineering 110, 190–199. Thoumine, O., Ott, A., 1997. Time scale dependent viscoelastic and contractile regimes in fibroblasts probed by microplate manipulation. Journal of cell science 110, 2109–2116. Valberg, P.A., Albertini, D.F., 1985. Cytoplasmic motions, rheology, and structure probed by a novel magnetic particle method. Journal of Cell Biology 101, 130–140. Waugh, R., Evans, E.A., 1976. Viscoelastic properties of erythrocyte membranes of different vertebrate animals. Microvascular Research 12, 291–304. Weisenhorn, A.L., Khorsandi, M., Kasas, S., Gotzos, V., Butt, H.J., 1993. Deformation and height anomaly of soft surfaces studied with an AFM. Nanotechnology 4, 106–113. Wu, Z.Z., Zhang, G., Long, M., Wang, H.B., Song, G.B., Cai, S.X., 2000. Comparison of the viscoelastic properties of normal hepatocytes and hepatocellular carcinoma cells under cytoskeletal perturbation. Biorheology 37, 279–290. Yamada, S., Wirtz, D., Kuo, S.C., 2000. Mechanics of living cells measured by laser tracking microrheology. Biophysical Journal 78, 1736–1747.