Eur Biophys J DOI 10.1007/s00249-014-1000-y
ORIGINAL PAPER
Nano‑characterization of two closely related melanoma cell lines with different metastatic potential Justyna Gostek · Szymon Prauzner‑Bechcicki · Benedikt Nimmervoll · Katrin Mayr · Joanna Pabijan · Peter Hinterdorfer · Lilia A. Chtcheglova · Małgorzata Lekka
Received: 20 August 2014 / Revised: 31 October 2014 / Accepted: 13 November 2014 © European Biophysical Societies’ Association 2014
Abstract Cutaneous malignant melanoma is one of the most lethal types of skin cancer. Its progression passes through several steps, leading to the appearance of a new population of cells with aggressive biological potential. Here, we focused on the nano-characterization of two different melanoma cell lines with similar morphological appearance but different metastatic potential, namely, WM115 from vertical growth phase (VGP) and WM266-4 derived from metastasis to skin. The first cell line represents cells that progressed to the VGP, while the WM266-4 cell line denotes cells from the metastasis to skin. Exploring with a combination of atomic force and fluorescence microscopes, our goal was to identify cell surface characteristics in both cell lines that may determine differences in the cellular nano-mechanical properties. Cell elasticity was found to be affected by the presence of F-actin-based flexible ridges, rich in F-actin co-localized with β1 integrins in L. A. Chtcheglova and M. Lekka contributed equally to this work. Electronic supplementary material The online version of this article (doi:10.1007/s00249-014-1000-y) contains supplementary material, which is available to authorized users. J. Gostek · S. Prauzner‑Bechcicki · M. Lekka (*) The Henryk Niewodniczan´ski Institute of Nuclear Physics, Polish Academy of Sciences, Radzikowskiego 152, 31‑342 Kraków, Poland e-mail: [email protected] B. Nimmervoll · K. Mayr · J. Pabijan · P. Hinterdorfer · L. A. Chtcheglova Center for Advanced Bioanalysis GmbH (CBL), Gruberstrasse 40‑42, 4020 Linz, Austria P. Hinterdorfer Institute for Biophysics, University of Linz, Gruberstrasse 40‑42, 4020 Linz, Austria
the studied cell lines. These results point out how progressive changes in the surface structure of melanoma cells can affect their bionanomechanical properties. Keywords Melanoma cells · Cell elasticity · Combined AFM/fluorescence microscopy · Correlation between surface morphology and cellular stiffness
Introduction Cells contain a cytoskeletal network, which is not only the 3D support for a cell but also participates in various cellular processes (Mofrad 2009). It defines the cellular mechanical properties and participates in the signaling towards cell nucleus by various surface receptors (Bershadsky et al. 2003). It has already been shown that cancer cells differ from normal ones in many aspects, including the rate of growth, adhesion, and cell morphology. One of these characteristics is the mechanical resistance maintained by the cytoskeletal structure and its organization (Hall 2009). Numerous literature data highlight that the organization of the cytoskeleton is altered in tumor progression (Jordan and Wilson 1998; Makale 2009). Thus, the information gained on the correlation between cellular surface structures and nano-mechanical properties would help to identify new potential biomarkers important for the development of effective cancer therapies. The technological advances made over the last decades in the fields of nanotechnology, engineering, and medicine have improved the characterization of properties of an individual cell that is further exploited for better understanding of the pathological changes occurring within a single cell. Atomic force microscopy (AFM) has become a highly valuable tool in the studies of biological samples ranging from
13
single proteins to living cells (Hinterdorfer and Dufrene 2009; Seo and Jhe 2008). The cellular stiffness has been shown to be an excellent marker of oncogenic progression in various cancers (Lekka et al. 2012; Xu et al. 2012). Distinct properties of cells from the early and late stages of tumor progression may become an indicator of malignant transformation and they could be used not only in diagnosis but also in the treatment and therapy (Brunner et al. 2009). Therefore, cell biomechanics studies in conjunction with analysis of its structure are essential for the improvement of advanced diagnostic methods and the development of new drugs. The incidence of cutaneous malignant melanoma in Caucasian populations has increased dramatically over the past 10 years. It has been classified into well-described clinical stages, ranging from lesions considered as benign to highly metastatic ones. Firstly, melanocytes can progress to a flat tumor growing horizontally (radial growth phase, RGP) followed by the acquisition of the capacity to invade the skin (vertical growth phase, VGP), ultimately leading to metastasis (Santiago-Walker et al. 2009). The transition from RGP to VGP is a biologically critical step accompanied by a variety of changes distinguishing between cells from RGP and VGP phases. The VGP cells change their phenotype, for example BRAF and NRAS mutations (Wellbrock et al. 2008; Kwong et al. 2012) towards more invasive one leading in consequence to further metastatic dissemination. The VGP cells express various cell surface molecules that are, later on, involved in metastasis. These receptors are mainly derived from a large family of heterodimeric molecules referred as the integrins that play key roles in the regulation of tumor cell migration and survival (Hynes 1992; Nikkola et al. 2004). Over 60 % of melanoma cases have been recorded with tumor thickness below 1 mm that, frequently, is considered to be non-metastatic melanoma (Criscione and Weinstock 2010). However, the increasing incidence of thin metastasizing melanoma (King et al. 2000) urges the researchers to determine other significant criteria and novel melanoma characteristics, both molecular and mechanical ones that are needed to improve the classifications of melanomas. Little is known about the nano-mechanical properties of melanoma cells, especially the correlation between their deformability and metastatic potential. Very early work, dating from 1988, quantified the B16 melanoma deformability as cell filterability by determining a percentage of cells traversing a 10 μm diameter pore at constant pressure as a function of time. With this system, it was demonstrated that cell deformability was related to metastatic potential, i.e. for more invasive melanoma cells, larger filterability was observed (Ochalek et al. 1988). In other work, dating from 2012, the same melanoma cell line, B16, was studied, but the results showed that cells with higher metastatic potential were more rigid (Watanabe et al. 2012).
13
Eur Biophys J
In our studies, we have focused on the nano-characterization of two closely related human melanoma cell lines (WM115 and WM266-4) originating from the same patient. WM115 cells were derived from the primary tumor in vertical growth phase (VGP), whereas WM266-4 cells were established from a metastasis to the skin. Our interest was to find a correlation between surface morphology and mechanical properties of melanoma cells. Therefore, the organization of cytoskeleton, i.e. F-actin filaments, was investigated by both fluorescence microscopy and atomic force microscopy (AFM). The apparent softness of the metastatic WM226-4 cells in comparison to the VGP cells (WM115) was correlated with the presence of actinbased flexible ridges on their surface. The combination of these features (or parameters) enabled us to clearly identify WM266-4 (metastatic) and WM115 (non-metastatic) cells despite their similar appearance detected with an optical microscope.
Materials and methods Cell cultures The WM115 and WM266-4 melanoma cells were grown in culture flasks in RPMI-1640 medium (Sigma-Aldrich) with 10 % heat-inactivated fetal bovine serum (FBS, SigmaAldrich) supplemented with a penicillin–streptomycin– neomycin mixture of antibiotics (Sigma-Aldrich) at 37 °C in 95 % air/5 % CO2 atmosphere. For AFM, epifluorescent and immunofluorescent cells were cultured following the corresponding protocols described in the Supplementary Materials, together with the experimental details on epifluorescence and confocal microscopies. Atomic force microscopes Topographical images of gently fixed cells were collected in MAC Mode using a commercial AFM set-up (AFM5500; Agilent Technologies, USA) and magnetically coated AFM tips (type 7; Agilent Technologies). The images of living cells in RPMI-1640 medium were collected at room temperature in contact mode using an AFM Xe120 (Park Systems, Korea). The measurements were accomplished using commercially available silicon nitride (Si3N4) cantilevers with the nominal spring constant of 0.01 N/m and tip radius of 50 nm (MLCT; Veeco, USA). In total, ~60 topographical images were recorded for WM115 and WM266-4 cells while a roughness was calculated for 30 cells. Force curves, i.e. dependence between the cantilever deflection and the relative scanner position, were recorded at the approach/retract velocity of 7 μm/s. The force curves were collected over a scan area ranging from 15 × 15 to
Eur Biophys J
20 × 20 µm2. Subsequently, a grid of 8 × 8 points was set over the scanned region. On average, 64 force curves for each single cell were automatically recorded. The distance between the probing locations (i.e. distance between two subsequent points of the grid) was from 1,875 to 2,500 nm. Such values were chosen to minimize the influence of cytoskeleton remodeling induced by the indentation. The total number of measured cells for each studied cell line varied between 15 and 25 cells. The measurements were repeated three times with fresh culture cells, medium, and new AFM cantilevers. The elasticity changes were statistically verified using Student’s t-test. Combined AFM/epifluorescence microscopy For combined AFM/epifluorescence measurements, cells were grown on glass bottom Petri dishes (MatTek, USA). Upon 30 % of confluence, cells were gently fixed with prewarmed 4 % formaldehyde in PBS buffer for 20 min and subsequently labeled with rhodamine phalloidin (1:500; Molecular Probes). Further studies were performed on a 6000 ILM AFM set-up (Agilent Technologies) integrated with an inverted optical microscope Zeiss Observer.D1 (Zeiss, Germany). Firstly, AFM topographical images of melanoma cells were collected in contact mode using a silicon nitride cantilever with nominal spring constant of 10 pN/nm (MLCT; Bruker, CA, USA). Fluorescent images were then acquired using an Oligochrome light source lamp (Till Photonics, Germany), Cy3 HC filter set (Semrock, USA; excitation/bandwidth: 531/40 nm and emission/ bandwidth: 593/40 nm), and a ×63 (oil immersion) objective (Zeiss) and finally captured by a Hamamatsu digital camera ORCA-Flash 2.8. Each AFM/fluorescence experiment was obtained at least in duplicate using independent cell preparations. Cell elasticity data analysis The determination of the elasticity modulus, characterizing the cell stiffness, is based on the force–indentation curve that reflects the interaction between the AFM tip and the investigated surface causing cantilever deflection (Lekka et al. 1999). The deflection depends on the applied force and the compliance of the investigated sample. If a rigid material (e.g., silicon or glass) is investigated, the deflection reflects the position of the sample. This is represented by a straight-sloped line. For soft samples, like cells, cantilever deflection is much smaller and the resulting force curve has a non-linear character. The difference between these curves determines the deformation of the sample surface. The relative value of the Young’s modulus is determined by fitting the Hertz–Sneddon model to the obtained force versus indentation curves, following the method described
previously (Lekka et al. 2001). The average value of the Young’s modulus was determined by fitting the Gaussian distribution to the histogram of the values obtained for each curve.
Results and discussion Combined topography AFM/epifluorescence microscopy imaging Firstly, we investigated the topography of single cells at large scan areas. Figure 1 represents the typical AFM images of gently fixed cells recorded using MAC mode. For this, melanoma cells were fixed with 4 % paraformaldehyde (PFA; Sigma) dissolved in Hank’s buffered salt solution (HBSS; Sigma) at 37 °C for at least 2 h. To verify whether the cell surface structure remains the same after fixation, the imaging of living cells (WM115 and WM266-4) was also carried out in contact mode (Figure S1). Independently of the cell type (WM115 or WM266-4), the shape of a typical single cell was very similar, showing a cell nucleus in a central region of the triangular-shaped cell with visible filamentous structure. The literature data (Defilippi et al. 1999; Braet et al. 2001) enabled us to assume that filaments observed in the AFM images were F-actin filaments. To investigate cell landscape at nanoscale (e.g., the organization of F-actin filaments beneath cell membrane), AFM images (256 × 256 pixels) on smaller scan areas of 15 × 15 μm2 (living cells; Figure S1B, D) and about 9 × 9 µm2 (gently fixed cells; Fig. 1c, d) were performed. A clear difference was observed in the arrangement of filaments lying directly beneath the cell membrane. In the case of WM115 cells (Fig. 1a, b), the observed cytoskeleton organization was characterized by the presence of long bundles of filaments interspersed with much shorter ones. Interestingly, such aligned organization of fibers disappeared for cells derived from metastatic cells (WM266-4) (Fig. 1c, d). The WM266-4 cell surface was found to be very rough and decorated with numerous dorsal “ruffles” or “ridges”. These ridges appeared to be flexible as they were easy to displace in the direction of a scan. Similar structures have also been described for other cells, such as skin fibroblasts and liver endothelial cells (Braet et al. 2001). To visualize the morphological organization of F-actin filaments in melanoma cells, and to identify F-actin on the cell topography images, we have performed correlated AFM (e.g., topography images) and epifluorescence measurements on the AFM set-up integrated with the inverted light microscope (El-Kirat-Chatel and Dufrêne 2012). This set-up enables us to use optical images to conduct precisely AFM imaging (Fig. 2). Bright-field (phase contrast)
13
Eur Biophys J
Fig. 1 Typical AFM topographical (heights) images of gently fixed melanoma cells (WM115 (a) 25 × 25 µm2 and WM266-4 (b) 29 × 29 µm2) and their fragments (about 9 × 9 µm2; 256 × 256 pixels, WM115 (c) and WM266-4 (d))
Fig. 2 Optically guided AFM measurements on cells. a Brightfield (phase contrast) image of fixed WM115 cells obtained with a ×20 objective. b Topography AFM image (deflection) of WM115 cell in the selected area (dashed square in a). The scan size was
13
30 × 30 µm2 with the height scale ranging from 0 to ~2 µm. c Overlay of optical image (phase contrast, taken with a ×63 objective) with AFM topographical image
Eur Biophys J
Fig. 3 Correlated topography AFM imaging and epifluorescence microscopy obtained for a–e WM115 and a′–e′ WM266-4 cells. Scale bars (a, b and a′, b′) 20 µm, (c–e and c′–e′) 5 µm. The F-actin was stained with rhodamine-phalloidin
optical images were first collected with a ×20 objective (Fig. 2a). Subsequently, AFM topographical views (simultaneous height and deflection images) of melanoma cells were recorded in contact mode (Fig. 2b). AFM topographical images recorded in these conditions were found to be very similar to those recorded previously, and illustrated in Fig. 1. After changing the optical objective (×63), an optical focus was adjusted to the glass surface and a series of epiluorescence images at different focus planes (from glass surface to the top of a cell) were collected (see Supplementary Figures S2A and S2B). Finally, the AFM images of WM115 and WM266-4 cells were correlated with optical/ fluorescent images, as shown in Fig. 3. More scrupulous image analysis was further performed using a FIJI platform (fluorescence background subtraction, image overlay). As shown in Fig. 3c, c′, the aligned filaments observed on the surface of WM115 cells (Fig. 3c) and the flexible ridges seen on the surface of WM266-4 (Fig. 3c′) were properly attributed to the F-actin identified in the corresponding fluorescence images. The F-actin was stained using phalloidin
labelled with rhodamine. Fluorescent images show nicely organized filaments in WM115 cells (Fig. 3d) while their organization was disordered in WM266-4 ones (Fig. 3d′). Morphology of cells in low and high cellular densities To follow the organization of F-actin fibers at a single cell level, and in conditions of semi-confluent monolayer, fluorescent images were recorded for both melanoma cell lines (WM115 and WM266-4) after 3 days of growing at two distinct densities. The confocal images illustrated two groups of F-actin organization either as ventral stress fibers or short F-actin microfilaments constituting a cortical mesh (Supplementary Figure S2). To quantify morphological properties, surface areas of both whole cells and nuclei were determined. They served as parameters to calculate the N/C ratio (i.e. a nucleus surface area divided by a cytoplasm surface area) that might be used as a marker to diagnose melanoma similarly to the cervical squamous epithelium (Walker et al. 2003). Melanoma cells cultured at higher density have
13
Eur Biophys J
Fig. 4 The Young’s modulus values (expressed as a mean ± standard deviation), obtained for living melanoma cells grown at low and high density. Results obtained for indentation depth of 50 nm (a), 200 nm
(b), and 500 nm (c) are shown, accompanied by their verification using Student’s t-test
displayed occasional partial overlap of cells (such effect was also visible for longer culture time; see Supplementary Table S1). The contribution of overlapping effect was mainly observed at higher cellular density with overall level of ~10–15 % of the covered surface. The surface area dropped significantly as the density of cells increased (from ~1,630 to ~1,130 µm2 and from ~1,500 to ~700 µm2, for WM115 and WM266-4 cells, respectively; see Supplementary Table S2), showing simultaneously changes in the area of contact between cell surface and substrate.
indentations (above 200 nm), contain more information about the organization of F-actin filaments in respect to cell membrane and cellular density. Thus, considering the cellular morphology based on AFM and fluorescence results, we hypothesize here that the elastic modulus of the studied melanoma cells is a consequence of the presence/absence of flexible ridges on a surface of melanoma cells. The presence of flexible ridges in WM266-4 melanoma cells prevents the long actin filaments from being squeezed by the AFM probe, resulting in lower Young’s modulus. As expected, the effect was independent of the indentation depth (Fig. 4b, c). The smoother surface of the WM115 cells allows easier access to F-actin filaments for the indenting AFM probe. Thus, at the same indentation depth, the WM155 cells appeared to be more rigid than those of WM266-4.
Nano‑mechanics of melanoma cells Both AFM and fluorescence images have illustrated various cytoskeletal organizations of WM115 and WM226-4 cells (Figs. 1, 2, 3, Figures S1, S2, S6). The degree of observed alterations can be quantified using the Young’s modulus value (describing the elastic properties of a cell). Our results have also demonstrated that cell stiffness is an indentation depth-dependent function (Fig. 4), which should be carefully considered when a comparison between various cell lines is performed (Lekka et al. 2012). During indentation, the probing AFM tip meets several structures, starting from glycocalix, cell membrane, actin cortex, and deeper parts of a cell. Therefore, for small indentation depths, the mechanical response will be dominated by cell membrane tension and the superficial layer of F-actin filaments. More profound indentations reflect the contributions from deeper layers, including the cell nucleus. Therefore, for the indentation depth equal to 50 nm (Fig. 4a), the Young’s modulus revealed large variations of elastic modulus that contain the overlapping moduli with dominating contributions from cell membrane, from F-actin filaments, or both. The obtained similar elastic moduli for both WM115 and WM266-4 cells, independently of cellular density, revealed the stiffness of the same materials, i.e. F-actin filaments and membrane, composing the superficial layers of the cell. Larger
13
F‑actin and β1‑integrin localization using confocal fluorescence microscopy The important issue concerns whether the increased motility of metastatic cells is associated with the expression of β1 integrins which can indicate more invasive states. The expression of integrins can change in malignant transformation. Despite their important role in cancer cell behavior, the merit of integrins as prognostic markers is unclear (see Supplementary section “The expression of β1 integrins in melanoma cells”). For both studied melanoma cells, a western blot was carried out on cell extracts (Figure S5A), and fluorescence staining with respective antibodies was performed in order to localize both F-actin and β1-integrins within the studied individual cells (Figures S5B and S5C, respectively). The presence of stress fibers (F-actin filaments) was observed for WM115 cells which led to their larger rigidity. At the same indentation depths observed for WM266-4 cells, the presence of flexible ridges colocalized with F-actin filaments is revealed in the smaller Young’s modulus values.
Eur Biophys J
Conclusion The AFM topographical images revealed considerable difference between WM115 and WM266-4 cells that were not visible under an optical microscope due to its low resolution (~200–300 nm) as compared to the AFM (several nm). The cytoskeleton of WM115 cells was more organized (cortical filamentous structure), whereas WM266-4 cells were rougher with the presence of characteristic flexible ridges (or ruffles). Since the ridges (or ruffles) were constantly detected only on WM266-4 cells, we assume that their presence might be associated with the melanoma progression. Moreover, β1 integrin was found to co-localize with F-actin, also present in flexible ridges. We have demonstrated that the combination of nonoptical atomic force microscopy and optical fluorescence microscopy enables the identification of cellular structures and nano-mechanical properties of melanoma cells. The obtained results show that larger deformability of metastatic WM266-4 melanoma cells correlates with the presence of flexible ridges which are not detectable by an optical microscope. Single melanoma cells start to be clearly identified by both the detailed images of their surface at nm scale and by their nano-mechanical properties. Our findings may help to understand the role of nano-mechanical properties in melanoma development and, furthermore, point out that the altered organization of F-actin filaments is not always responsible for cell softness. Recent studies have revealed significant differences in the nano-mechanical properties of benign and cancerous cells that were attributed to alterations in F-actin filament organization. However, the larger deformability of cancer cells is not always caused by the reorganization of actin cytoskeleton. Here, we have demonstrated that the elasticity of melanoma cells is correlated with the presence of F-actinbased flexible ridges, typical only for metastatic melanoma cells. These results may help to understand the role of nanomechanical properties in melanoma development. Acknowledgments This work was carried out within the collaboration between Austria–Poland scientific technological cooperation (AUT: PL03/2011; PL: 8507/R11/R12), financially supported by the project NCN DEC-2011/01/M/ST3/00711 (PL) and in the frame of Regio13 by the European Regional Development Fund (EFRE) and the state of Upper Austria (to L.A.C., K.M. and P.H.). Both institutions are also grateful to the COST Action TD1002 (AFM4NanoBioMed).
References Bershadsky AD, Balaban NQ, Geiger B (2003) Adhesion-dependent cell mechanosensitivity. Annu Rev Cell Dev Biol 19:677–695 Braet F, de Zanger R, Seynaeve C, Baekland M, Wisse E (2001) A comparative atomic force microscopy study on living skin fibroblasts and liver endothelial cells. J Electron Microsc 50:283–290 Brunner C, Niendorf A, Käs J (2009) Passive and active single-cell biomechanics: a new perspective in cancer diagnosis. Soft Matter 5:2171–2178
Criscione VD, Weinstock MA (2010) Melanoma thickness trends in the United States, 1988–2006. J Invest Dermatol 130:793–797 Defilippi P, Olivo C, Venturino M, Dolce L, Silengo L, Tarone G (1999) Actin cytoskeleton organization in response to integrinmediated adhesion. Microsc Res Tech 47:67–78 El-Kirat-Chatel S, Dufrêne YF (2012) Nanoscale imaging of the candida–macrophage interaction using correlated fluorescenceatomic force microscopy. ACS Nano 6:10792–10799 Hall A (2009) The cytoskeleton and cancer. Cancer Metastasis Rev 28:5–14 Hinterdorfer P, Dufrene Y (2009) Detection and localization of single molecular recognition events using atomic force microscopy. Nat Methods 3:347–355 Hynes RO (1992) Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69:11–25 Jordan M, Wilson L (1998) Microtubules and actin filaments: dynamic targets for cancer chemotherapy. Curr Opin Cell Biol 10:123–130 King R, Googe PB, Mihm MC (2000) Thin melanomas. Clin Lab Med 20:713–729 Kwong LN, Costello JC, Liu H, Jiang S, Helms TL, Langsdorf AE, Jakubosky D, Genovese G, Muller FL, Jeong JH, Bender RP, Chu GC, Flaherty KT, Wargo JA, Collins JJ, Chin L (2012) Oncogenic NRAS signaling differentially regulates survival and proliferation in melanoma. Nat Methods 18:1503–1510 Lekka M, Laidler P, Gil D, Lekki J, Stachura Z, Hrynkiewicz AZ (1999) Elasticity of normal and cancerous human bladder cells studied by scanning force microscopy. Eur Biophys J 28:312–316 Lekka M, Laidler P, Ignacak J, Labedz M, Lekki J, Struszczyk H, Stachura Z, Hrynkiewicz AZ (2001) The effect of chitosan on stiffness and glycolytic activity of human bladder cells. Biochim Biophys Acta 1540:127–136 Lekka M, Gil D, Pogoda K, Dulin´ska-Litewka J, Jach R, Gostek J, Klymenko O, Prauzner-Bechcicki S, Stachura Z, WiltowskaZuber J, Okon´ K, Laidler P (2012) Cancer cell detection in tissue sections using AFM. Arch Biochem Biophys 518:151–156 Makale M (2009) Cellular mechanobiology and cancer metastasis. Birth Defects Res C 81:329–343 Mofrad MRK (2009) Rheology of the Cytoskeleton. Annu Rev Fluid Mech 41:433–453 Nikkola J, Vihinen P, Vlaykova T, Hahka-Kemppinen M, Heino J, Pyrhönen S (2004) Integrin chains [beta]1 and [alpha]v as prognostic factors in human metastatic melanoma. Melanoma Res 14:29–37 Ochalek T, Nordt FJ, Tullberg K, Burger MM (1988) Correlation between cell deformability and metastatic potential in B16-F1 melanoma cell variants. Cancer Res 48:5124–5128 Santiago-Walker A, Li L, Haass N, Herlyn M (2009) Melanocytes: from morphology to application. Skin Pharmacol Physiol 22:114–121 Seo Y, Jhe W (2008) Atomic force microscopy and spectroscopy. Rep Prog Phys 71:016101 Walker DC, Brown BH, Blackett AD, Tidy J, Smallwood RH (2003) A study of the morphological parameters of cervical squamous epithelium. Physiol Meas 24:121–135 Watanabe T, Kuramochi H, Takahashi A, Imai K, Katsuta N, Nakayama T, Fujiki H, Suganuma M (2012) Higher cell stiffness indicating lower metastatic potential on B16 melanoma cell variants and in (−)-epigallocatehin gallate-treated cells. J Cancer Res Clin Oncol 138:859–866 Wellbrock C, Rana S, Paterson H, Pickersgill H, Brummelkamp T, Marais R (2008) Oncogenic BRAF regulates melanoma proliferation through the lineage specific factor MITF. PLoS ONE 3:e2734 Xu W, Mezencev R, Kim B, Wand L, McDonald J, Sulchek T (2012) Cell stiffness is a biomarker of the metastatic potential of ovarian cancer cells. PLoS ONE 7:e46609
13
ORIGINAL PAPER
Nano‑characterization of two closely related melanoma cell lines with different metastatic potential Justyna Gostek · Szymon Prauzner‑Bechcicki · Benedikt Nimmervoll · Katrin Mayr · Joanna Pabijan · Peter Hinterdorfer · Lilia A. Chtcheglova · Małgorzata Lekka
Received: 20 August 2014 / Revised: 31 October 2014 / Accepted: 13 November 2014 © European Biophysical Societies’ Association 2014
Abstract Cutaneous malignant melanoma is one of the most lethal types of skin cancer. Its progression passes through several steps, leading to the appearance of a new population of cells with aggressive biological potential. Here, we focused on the nano-characterization of two different melanoma cell lines with similar morphological appearance but different metastatic potential, namely, WM115 from vertical growth phase (VGP) and WM266-4 derived from metastasis to skin. The first cell line represents cells that progressed to the VGP, while the WM266-4 cell line denotes cells from the metastasis to skin. Exploring with a combination of atomic force and fluorescence microscopes, our goal was to identify cell surface characteristics in both cell lines that may determine differences in the cellular nano-mechanical properties. Cell elasticity was found to be affected by the presence of F-actin-based flexible ridges, rich in F-actin co-localized with β1 integrins in L. A. Chtcheglova and M. Lekka contributed equally to this work. Electronic supplementary material The online version of this article (doi:10.1007/s00249-014-1000-y) contains supplementary material, which is available to authorized users. J. Gostek · S. Prauzner‑Bechcicki · M. Lekka (*) The Henryk Niewodniczan´ski Institute of Nuclear Physics, Polish Academy of Sciences, Radzikowskiego 152, 31‑342 Kraków, Poland e-mail: [email protected] B. Nimmervoll · K. Mayr · J. Pabijan · P. Hinterdorfer · L. A. Chtcheglova Center for Advanced Bioanalysis GmbH (CBL), Gruberstrasse 40‑42, 4020 Linz, Austria P. Hinterdorfer Institute for Biophysics, University of Linz, Gruberstrasse 40‑42, 4020 Linz, Austria
the studied cell lines. These results point out how progressive changes in the surface structure of melanoma cells can affect their bionanomechanical properties. Keywords Melanoma cells · Cell elasticity · Combined AFM/fluorescence microscopy · Correlation between surface morphology and cellular stiffness
Introduction Cells contain a cytoskeletal network, which is not only the 3D support for a cell but also participates in various cellular processes (Mofrad 2009). It defines the cellular mechanical properties and participates in the signaling towards cell nucleus by various surface receptors (Bershadsky et al. 2003). It has already been shown that cancer cells differ from normal ones in many aspects, including the rate of growth, adhesion, and cell morphology. One of these characteristics is the mechanical resistance maintained by the cytoskeletal structure and its organization (Hall 2009). Numerous literature data highlight that the organization of the cytoskeleton is altered in tumor progression (Jordan and Wilson 1998; Makale 2009). Thus, the information gained on the correlation between cellular surface structures and nano-mechanical properties would help to identify new potential biomarkers important for the development of effective cancer therapies. The technological advances made over the last decades in the fields of nanotechnology, engineering, and medicine have improved the characterization of properties of an individual cell that is further exploited for better understanding of the pathological changes occurring within a single cell. Atomic force microscopy (AFM) has become a highly valuable tool in the studies of biological samples ranging from
13
single proteins to living cells (Hinterdorfer and Dufrene 2009; Seo and Jhe 2008). The cellular stiffness has been shown to be an excellent marker of oncogenic progression in various cancers (Lekka et al. 2012; Xu et al. 2012). Distinct properties of cells from the early and late stages of tumor progression may become an indicator of malignant transformation and they could be used not only in diagnosis but also in the treatment and therapy (Brunner et al. 2009). Therefore, cell biomechanics studies in conjunction with analysis of its structure are essential for the improvement of advanced diagnostic methods and the development of new drugs. The incidence of cutaneous malignant melanoma in Caucasian populations has increased dramatically over the past 10 years. It has been classified into well-described clinical stages, ranging from lesions considered as benign to highly metastatic ones. Firstly, melanocytes can progress to a flat tumor growing horizontally (radial growth phase, RGP) followed by the acquisition of the capacity to invade the skin (vertical growth phase, VGP), ultimately leading to metastasis (Santiago-Walker et al. 2009). The transition from RGP to VGP is a biologically critical step accompanied by a variety of changes distinguishing between cells from RGP and VGP phases. The VGP cells change their phenotype, for example BRAF and NRAS mutations (Wellbrock et al. 2008; Kwong et al. 2012) towards more invasive one leading in consequence to further metastatic dissemination. The VGP cells express various cell surface molecules that are, later on, involved in metastasis. These receptors are mainly derived from a large family of heterodimeric molecules referred as the integrins that play key roles in the regulation of tumor cell migration and survival (Hynes 1992; Nikkola et al. 2004). Over 60 % of melanoma cases have been recorded with tumor thickness below 1 mm that, frequently, is considered to be non-metastatic melanoma (Criscione and Weinstock 2010). However, the increasing incidence of thin metastasizing melanoma (King et al. 2000) urges the researchers to determine other significant criteria and novel melanoma characteristics, both molecular and mechanical ones that are needed to improve the classifications of melanomas. Little is known about the nano-mechanical properties of melanoma cells, especially the correlation between their deformability and metastatic potential. Very early work, dating from 1988, quantified the B16 melanoma deformability as cell filterability by determining a percentage of cells traversing a 10 μm diameter pore at constant pressure as a function of time. With this system, it was demonstrated that cell deformability was related to metastatic potential, i.e. for more invasive melanoma cells, larger filterability was observed (Ochalek et al. 1988). In other work, dating from 2012, the same melanoma cell line, B16, was studied, but the results showed that cells with higher metastatic potential were more rigid (Watanabe et al. 2012).
13
Eur Biophys J
In our studies, we have focused on the nano-characterization of two closely related human melanoma cell lines (WM115 and WM266-4) originating from the same patient. WM115 cells were derived from the primary tumor in vertical growth phase (VGP), whereas WM266-4 cells were established from a metastasis to the skin. Our interest was to find a correlation between surface morphology and mechanical properties of melanoma cells. Therefore, the organization of cytoskeleton, i.e. F-actin filaments, was investigated by both fluorescence microscopy and atomic force microscopy (AFM). The apparent softness of the metastatic WM226-4 cells in comparison to the VGP cells (WM115) was correlated with the presence of actinbased flexible ridges on their surface. The combination of these features (or parameters) enabled us to clearly identify WM266-4 (metastatic) and WM115 (non-metastatic) cells despite their similar appearance detected with an optical microscope.
Materials and methods Cell cultures The WM115 and WM266-4 melanoma cells were grown in culture flasks in RPMI-1640 medium (Sigma-Aldrich) with 10 % heat-inactivated fetal bovine serum (FBS, SigmaAldrich) supplemented with a penicillin–streptomycin– neomycin mixture of antibiotics (Sigma-Aldrich) at 37 °C in 95 % air/5 % CO2 atmosphere. For AFM, epifluorescent and immunofluorescent cells were cultured following the corresponding protocols described in the Supplementary Materials, together with the experimental details on epifluorescence and confocal microscopies. Atomic force microscopes Topographical images of gently fixed cells were collected in MAC Mode using a commercial AFM set-up (AFM5500; Agilent Technologies, USA) and magnetically coated AFM tips (type 7; Agilent Technologies). The images of living cells in RPMI-1640 medium were collected at room temperature in contact mode using an AFM Xe120 (Park Systems, Korea). The measurements were accomplished using commercially available silicon nitride (Si3N4) cantilevers with the nominal spring constant of 0.01 N/m and tip radius of 50 nm (MLCT; Veeco, USA). In total, ~60 topographical images were recorded for WM115 and WM266-4 cells while a roughness was calculated for 30 cells. Force curves, i.e. dependence between the cantilever deflection and the relative scanner position, were recorded at the approach/retract velocity of 7 μm/s. The force curves were collected over a scan area ranging from 15 × 15 to
Eur Biophys J
20 × 20 µm2. Subsequently, a grid of 8 × 8 points was set over the scanned region. On average, 64 force curves for each single cell were automatically recorded. The distance between the probing locations (i.e. distance between two subsequent points of the grid) was from 1,875 to 2,500 nm. Such values were chosen to minimize the influence of cytoskeleton remodeling induced by the indentation. The total number of measured cells for each studied cell line varied between 15 and 25 cells. The measurements were repeated three times with fresh culture cells, medium, and new AFM cantilevers. The elasticity changes were statistically verified using Student’s t-test. Combined AFM/epifluorescence microscopy For combined AFM/epifluorescence measurements, cells were grown on glass bottom Petri dishes (MatTek, USA). Upon 30 % of confluence, cells were gently fixed with prewarmed 4 % formaldehyde in PBS buffer for 20 min and subsequently labeled with rhodamine phalloidin (1:500; Molecular Probes). Further studies were performed on a 6000 ILM AFM set-up (Agilent Technologies) integrated with an inverted optical microscope Zeiss Observer.D1 (Zeiss, Germany). Firstly, AFM topographical images of melanoma cells were collected in contact mode using a silicon nitride cantilever with nominal spring constant of 10 pN/nm (MLCT; Bruker, CA, USA). Fluorescent images were then acquired using an Oligochrome light source lamp (Till Photonics, Germany), Cy3 HC filter set (Semrock, USA; excitation/bandwidth: 531/40 nm and emission/ bandwidth: 593/40 nm), and a ×63 (oil immersion) objective (Zeiss) and finally captured by a Hamamatsu digital camera ORCA-Flash 2.8. Each AFM/fluorescence experiment was obtained at least in duplicate using independent cell preparations. Cell elasticity data analysis The determination of the elasticity modulus, characterizing the cell stiffness, is based on the force–indentation curve that reflects the interaction between the AFM tip and the investigated surface causing cantilever deflection (Lekka et al. 1999). The deflection depends on the applied force and the compliance of the investigated sample. If a rigid material (e.g., silicon or glass) is investigated, the deflection reflects the position of the sample. This is represented by a straight-sloped line. For soft samples, like cells, cantilever deflection is much smaller and the resulting force curve has a non-linear character. The difference between these curves determines the deformation of the sample surface. The relative value of the Young’s modulus is determined by fitting the Hertz–Sneddon model to the obtained force versus indentation curves, following the method described
previously (Lekka et al. 2001). The average value of the Young’s modulus was determined by fitting the Gaussian distribution to the histogram of the values obtained for each curve.
Results and discussion Combined topography AFM/epifluorescence microscopy imaging Firstly, we investigated the topography of single cells at large scan areas. Figure 1 represents the typical AFM images of gently fixed cells recorded using MAC mode. For this, melanoma cells were fixed with 4 % paraformaldehyde (PFA; Sigma) dissolved in Hank’s buffered salt solution (HBSS; Sigma) at 37 °C for at least 2 h. To verify whether the cell surface structure remains the same after fixation, the imaging of living cells (WM115 and WM266-4) was also carried out in contact mode (Figure S1). Independently of the cell type (WM115 or WM266-4), the shape of a typical single cell was very similar, showing a cell nucleus in a central region of the triangular-shaped cell with visible filamentous structure. The literature data (Defilippi et al. 1999; Braet et al. 2001) enabled us to assume that filaments observed in the AFM images were F-actin filaments. To investigate cell landscape at nanoscale (e.g., the organization of F-actin filaments beneath cell membrane), AFM images (256 × 256 pixels) on smaller scan areas of 15 × 15 μm2 (living cells; Figure S1B, D) and about 9 × 9 µm2 (gently fixed cells; Fig. 1c, d) were performed. A clear difference was observed in the arrangement of filaments lying directly beneath the cell membrane. In the case of WM115 cells (Fig. 1a, b), the observed cytoskeleton organization was characterized by the presence of long bundles of filaments interspersed with much shorter ones. Interestingly, such aligned organization of fibers disappeared for cells derived from metastatic cells (WM266-4) (Fig. 1c, d). The WM266-4 cell surface was found to be very rough and decorated with numerous dorsal “ruffles” or “ridges”. These ridges appeared to be flexible as they were easy to displace in the direction of a scan. Similar structures have also been described for other cells, such as skin fibroblasts and liver endothelial cells (Braet et al. 2001). To visualize the morphological organization of F-actin filaments in melanoma cells, and to identify F-actin on the cell topography images, we have performed correlated AFM (e.g., topography images) and epifluorescence measurements on the AFM set-up integrated with the inverted light microscope (El-Kirat-Chatel and Dufrêne 2012). This set-up enables us to use optical images to conduct precisely AFM imaging (Fig. 2). Bright-field (phase contrast)
13
Eur Biophys J
Fig. 1 Typical AFM topographical (heights) images of gently fixed melanoma cells (WM115 (a) 25 × 25 µm2 and WM266-4 (b) 29 × 29 µm2) and their fragments (about 9 × 9 µm2; 256 × 256 pixels, WM115 (c) and WM266-4 (d))
Fig. 2 Optically guided AFM measurements on cells. a Brightfield (phase contrast) image of fixed WM115 cells obtained with a ×20 objective. b Topography AFM image (deflection) of WM115 cell in the selected area (dashed square in a). The scan size was
13
30 × 30 µm2 with the height scale ranging from 0 to ~2 µm. c Overlay of optical image (phase contrast, taken with a ×63 objective) with AFM topographical image
Eur Biophys J
Fig. 3 Correlated topography AFM imaging and epifluorescence microscopy obtained for a–e WM115 and a′–e′ WM266-4 cells. Scale bars (a, b and a′, b′) 20 µm, (c–e and c′–e′) 5 µm. The F-actin was stained with rhodamine-phalloidin
optical images were first collected with a ×20 objective (Fig. 2a). Subsequently, AFM topographical views (simultaneous height and deflection images) of melanoma cells were recorded in contact mode (Fig. 2b). AFM topographical images recorded in these conditions were found to be very similar to those recorded previously, and illustrated in Fig. 1. After changing the optical objective (×63), an optical focus was adjusted to the glass surface and a series of epiluorescence images at different focus planes (from glass surface to the top of a cell) were collected (see Supplementary Figures S2A and S2B). Finally, the AFM images of WM115 and WM266-4 cells were correlated with optical/ fluorescent images, as shown in Fig. 3. More scrupulous image analysis was further performed using a FIJI platform (fluorescence background subtraction, image overlay). As shown in Fig. 3c, c′, the aligned filaments observed on the surface of WM115 cells (Fig. 3c) and the flexible ridges seen on the surface of WM266-4 (Fig. 3c′) were properly attributed to the F-actin identified in the corresponding fluorescence images. The F-actin was stained using phalloidin
labelled with rhodamine. Fluorescent images show nicely organized filaments in WM115 cells (Fig. 3d) while their organization was disordered in WM266-4 ones (Fig. 3d′). Morphology of cells in low and high cellular densities To follow the organization of F-actin fibers at a single cell level, and in conditions of semi-confluent monolayer, fluorescent images were recorded for both melanoma cell lines (WM115 and WM266-4) after 3 days of growing at two distinct densities. The confocal images illustrated two groups of F-actin organization either as ventral stress fibers or short F-actin microfilaments constituting a cortical mesh (Supplementary Figure S2). To quantify morphological properties, surface areas of both whole cells and nuclei were determined. They served as parameters to calculate the N/C ratio (i.e. a nucleus surface area divided by a cytoplasm surface area) that might be used as a marker to diagnose melanoma similarly to the cervical squamous epithelium (Walker et al. 2003). Melanoma cells cultured at higher density have
13
Eur Biophys J
Fig. 4 The Young’s modulus values (expressed as a mean ± standard deviation), obtained for living melanoma cells grown at low and high density. Results obtained for indentation depth of 50 nm (a), 200 nm
(b), and 500 nm (c) are shown, accompanied by their verification using Student’s t-test
displayed occasional partial overlap of cells (such effect was also visible for longer culture time; see Supplementary Table S1). The contribution of overlapping effect was mainly observed at higher cellular density with overall level of ~10–15 % of the covered surface. The surface area dropped significantly as the density of cells increased (from ~1,630 to ~1,130 µm2 and from ~1,500 to ~700 µm2, for WM115 and WM266-4 cells, respectively; see Supplementary Table S2), showing simultaneously changes in the area of contact between cell surface and substrate.
indentations (above 200 nm), contain more information about the organization of F-actin filaments in respect to cell membrane and cellular density. Thus, considering the cellular morphology based on AFM and fluorescence results, we hypothesize here that the elastic modulus of the studied melanoma cells is a consequence of the presence/absence of flexible ridges on a surface of melanoma cells. The presence of flexible ridges in WM266-4 melanoma cells prevents the long actin filaments from being squeezed by the AFM probe, resulting in lower Young’s modulus. As expected, the effect was independent of the indentation depth (Fig. 4b, c). The smoother surface of the WM115 cells allows easier access to F-actin filaments for the indenting AFM probe. Thus, at the same indentation depth, the WM155 cells appeared to be more rigid than those of WM266-4.
Nano‑mechanics of melanoma cells Both AFM and fluorescence images have illustrated various cytoskeletal organizations of WM115 and WM226-4 cells (Figs. 1, 2, 3, Figures S1, S2, S6). The degree of observed alterations can be quantified using the Young’s modulus value (describing the elastic properties of a cell). Our results have also demonstrated that cell stiffness is an indentation depth-dependent function (Fig. 4), which should be carefully considered when a comparison between various cell lines is performed (Lekka et al. 2012). During indentation, the probing AFM tip meets several structures, starting from glycocalix, cell membrane, actin cortex, and deeper parts of a cell. Therefore, for small indentation depths, the mechanical response will be dominated by cell membrane tension and the superficial layer of F-actin filaments. More profound indentations reflect the contributions from deeper layers, including the cell nucleus. Therefore, for the indentation depth equal to 50 nm (Fig. 4a), the Young’s modulus revealed large variations of elastic modulus that contain the overlapping moduli with dominating contributions from cell membrane, from F-actin filaments, or both. The obtained similar elastic moduli for both WM115 and WM266-4 cells, independently of cellular density, revealed the stiffness of the same materials, i.e. F-actin filaments and membrane, composing the superficial layers of the cell. Larger
13
F‑actin and β1‑integrin localization using confocal fluorescence microscopy The important issue concerns whether the increased motility of metastatic cells is associated with the expression of β1 integrins which can indicate more invasive states. The expression of integrins can change in malignant transformation. Despite their important role in cancer cell behavior, the merit of integrins as prognostic markers is unclear (see Supplementary section “The expression of β1 integrins in melanoma cells”). For both studied melanoma cells, a western blot was carried out on cell extracts (Figure S5A), and fluorescence staining with respective antibodies was performed in order to localize both F-actin and β1-integrins within the studied individual cells (Figures S5B and S5C, respectively). The presence of stress fibers (F-actin filaments) was observed for WM115 cells which led to their larger rigidity. At the same indentation depths observed for WM266-4 cells, the presence of flexible ridges colocalized with F-actin filaments is revealed in the smaller Young’s modulus values.
Eur Biophys J
Conclusion The AFM topographical images revealed considerable difference between WM115 and WM266-4 cells that were not visible under an optical microscope due to its low resolution (~200–300 nm) as compared to the AFM (several nm). The cytoskeleton of WM115 cells was more organized (cortical filamentous structure), whereas WM266-4 cells were rougher with the presence of characteristic flexible ridges (or ruffles). Since the ridges (or ruffles) were constantly detected only on WM266-4 cells, we assume that their presence might be associated with the melanoma progression. Moreover, β1 integrin was found to co-localize with F-actin, also present in flexible ridges. We have demonstrated that the combination of nonoptical atomic force microscopy and optical fluorescence microscopy enables the identification of cellular structures and nano-mechanical properties of melanoma cells. The obtained results show that larger deformability of metastatic WM266-4 melanoma cells correlates with the presence of flexible ridges which are not detectable by an optical microscope. Single melanoma cells start to be clearly identified by both the detailed images of their surface at nm scale and by their nano-mechanical properties. Our findings may help to understand the role of nano-mechanical properties in melanoma development and, furthermore, point out that the altered organization of F-actin filaments is not always responsible for cell softness. Recent studies have revealed significant differences in the nano-mechanical properties of benign and cancerous cells that were attributed to alterations in F-actin filament organization. However, the larger deformability of cancer cells is not always caused by the reorganization of actin cytoskeleton. Here, we have demonstrated that the elasticity of melanoma cells is correlated with the presence of F-actinbased flexible ridges, typical only for metastatic melanoma cells. These results may help to understand the role of nanomechanical properties in melanoma development. Acknowledgments This work was carried out within the collaboration between Austria–Poland scientific technological cooperation (AUT: PL03/2011; PL: 8507/R11/R12), financially supported by the project NCN DEC-2011/01/M/ST3/00711 (PL) and in the frame of Regio13 by the European Regional Development Fund (EFRE) and the state of Upper Austria (to L.A.C., K.M. and P.H.). Both institutions are also grateful to the COST Action TD1002 (AFM4NanoBioMed).
References Bershadsky AD, Balaban NQ, Geiger B (2003) Adhesion-dependent cell mechanosensitivity. Annu Rev Cell Dev Biol 19:677–695 Braet F, de Zanger R, Seynaeve C, Baekland M, Wisse E (2001) A comparative atomic force microscopy study on living skin fibroblasts and liver endothelial cells. J Electron Microsc 50:283–290 Brunner C, Niendorf A, Käs J (2009) Passive and active single-cell biomechanics: a new perspective in cancer diagnosis. Soft Matter 5:2171–2178
Criscione VD, Weinstock MA (2010) Melanoma thickness trends in the United States, 1988–2006. J Invest Dermatol 130:793–797 Defilippi P, Olivo C, Venturino M, Dolce L, Silengo L, Tarone G (1999) Actin cytoskeleton organization in response to integrinmediated adhesion. Microsc Res Tech 47:67–78 El-Kirat-Chatel S, Dufrêne YF (2012) Nanoscale imaging of the candida–macrophage interaction using correlated fluorescenceatomic force microscopy. ACS Nano 6:10792–10799 Hall A (2009) The cytoskeleton and cancer. Cancer Metastasis Rev 28:5–14 Hinterdorfer P, Dufrene Y (2009) Detection and localization of single molecular recognition events using atomic force microscopy. Nat Methods 3:347–355 Hynes RO (1992) Integrins: versatility, modulation, and signaling in cell adhesion. Cell 69:11–25 Jordan M, Wilson L (1998) Microtubules and actin filaments: dynamic targets for cancer chemotherapy. Curr Opin Cell Biol 10:123–130 King R, Googe PB, Mihm MC (2000) Thin melanomas. Clin Lab Med 20:713–729 Kwong LN, Costello JC, Liu H, Jiang S, Helms TL, Langsdorf AE, Jakubosky D, Genovese G, Muller FL, Jeong JH, Bender RP, Chu GC, Flaherty KT, Wargo JA, Collins JJ, Chin L (2012) Oncogenic NRAS signaling differentially regulates survival and proliferation in melanoma. Nat Methods 18:1503–1510 Lekka M, Laidler P, Gil D, Lekki J, Stachura Z, Hrynkiewicz AZ (1999) Elasticity of normal and cancerous human bladder cells studied by scanning force microscopy. Eur Biophys J 28:312–316 Lekka M, Laidler P, Ignacak J, Labedz M, Lekki J, Struszczyk H, Stachura Z, Hrynkiewicz AZ (2001) The effect of chitosan on stiffness and glycolytic activity of human bladder cells. Biochim Biophys Acta 1540:127–136 Lekka M, Gil D, Pogoda K, Dulin´ska-Litewka J, Jach R, Gostek J, Klymenko O, Prauzner-Bechcicki S, Stachura Z, WiltowskaZuber J, Okon´ K, Laidler P (2012) Cancer cell detection in tissue sections using AFM. Arch Biochem Biophys 518:151–156 Makale M (2009) Cellular mechanobiology and cancer metastasis. Birth Defects Res C 81:329–343 Mofrad MRK (2009) Rheology of the Cytoskeleton. Annu Rev Fluid Mech 41:433–453 Nikkola J, Vihinen P, Vlaykova T, Hahka-Kemppinen M, Heino J, Pyrhönen S (2004) Integrin chains [beta]1 and [alpha]v as prognostic factors in human metastatic melanoma. Melanoma Res 14:29–37 Ochalek T, Nordt FJ, Tullberg K, Burger MM (1988) Correlation between cell deformability and metastatic potential in B16-F1 melanoma cell variants. Cancer Res 48:5124–5128 Santiago-Walker A, Li L, Haass N, Herlyn M (2009) Melanocytes: from morphology to application. Skin Pharmacol Physiol 22:114–121 Seo Y, Jhe W (2008) Atomic force microscopy and spectroscopy. Rep Prog Phys 71:016101 Walker DC, Brown BH, Blackett AD, Tidy J, Smallwood RH (2003) A study of the morphological parameters of cervical squamous epithelium. Physiol Meas 24:121–135 Watanabe T, Kuramochi H, Takahashi A, Imai K, Katsuta N, Nakayama T, Fujiki H, Suganuma M (2012) Higher cell stiffness indicating lower metastatic potential on B16 melanoma cell variants and in (−)-epigallocatehin gallate-treated cells. J Cancer Res Clin Oncol 138:859–866 Wellbrock C, Rana S, Paterson H, Pickersgill H, Brummelkamp T, Marais R (2008) Oncogenic BRAF regulates melanoma proliferation through the lineage specific factor MITF. PLoS ONE 3:e2734 Xu W, Mezencev R, Kim B, Wand L, McDonald J, Sulchek T (2012) Cell stiffness is a biomarker of the metastatic potential of ovarian cancer cells. PLoS ONE 7:e46609
13
