Biomed Microdevices (2015) 17:46 DOI 10.1007/s10544-015-9950-0
A soft, stretchable and conductive biointerface for cell mechanobiology Irene Bernardeschi & Francesco Greco & Gianni Ciofani & Attilio Marino & Virgilio Mattoli & Barbara Mazzolai & Lucia Beccai
# Springer Science+Business Media New York 2015
Abstract In mechanobiology the study of cell response to mechanical stimuli is fundamental, and the involved processes (i.e., mechanotransduction) need to be investigated by interfacing (mechanically and electrically) with the cells in dynamic and non-invasive natural-like conditions. In this work, we present a novel soft, stretchable and conductive biointerface that allows both cell mechanical stimulation and dynamic impedance recording. The biointerface stretchability and conductivity, jointly to the biocompatibility and transparency needed to perform cell culture studies, were obtained by exploiting the formation of wrinkles on the surface of a 90 nm thick conductive layer of poly(3,4ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) on a pre-stretched 130 μm thick poly(dimethylsiloxane) (PDMS) substrate. Cell adhesion and proliferation of SH-SY5Y human neuroblastoma cells were evaluated, and cell differentiation on the corrugated surface was assessed. We demonstrate how the biointerface remains conductive when applying uniaxial strain up to 10 %, and when cell culturing is performed. Finally, a reduction of about 30 % of the relative impedance variation signal was measured, with respect to the control, as a result of the mechanical stimulation of cells.
Electronic supplementary material The online version of this article (doi:10.1007/s10544-015-9950-0) contains supplementary material, which is available to authorized users. I. Bernardeschi : F. Greco : G. Ciofani : A. Marino : V. Mattoli : B. Mazzolai : L. Beccai (*) Center for Micro-BioRobotics, Istituto Italiano di Tecnologia, Viale Rinaldo Piaggio 34, 56025 Pontedera, PI, Italy e-mail: [email protected] I. Bernardeschi : A. Marino The Biorobotics Institute, Scuola Superiore Sant’Anna, Viale Rinaldo Piaggio 34, 56025 Pontedera, Italy
Keywords Stretchable biointerface . Impedance . Biohybrid . Surface wrinkling . PEDOT:PSS . Mechanotransduction
1 Introduction Living organisms, from plants and microorganisms to mammals, are mechanosensitive that is they can feel and react to external forces in order to interact with the environment where they live. Mechanobiology studies the effect of such forces on cells and, in particular, is focused on the understanding of mechanotransduction, a basic process by which cells can convert external mechanical stimuli into biochemical signals. Mechanical forces regulate cells biochemistry and the correct functioning of physiological processes and homeostasis are maintained through the mechanotransduction process (Ingber 2003; Jaalouk and Lammerding 2009; Orr et al. 2006). Indeed, mechanotransduction studies have a significant role in recognizing causes for some pathologies associated to an incorrect regulation of physiological processes (e.g., a defect in protein expression due to an wrong transduction process) (Coughlin et al. 2008; Heydemann and McNally 2007). However, it is mostly in sensing that mechanotransduction plays a main role; specifically, in hearing and touch, where external mechanical forces - produced by, e.g., sound waves, pressure or stretch - act on the dedicated sensing organs (i.e., outer and inner ear and skin), where they are detected and transformed into electrochemical signals. Investigating mechanotransduction requires the study of cell response to mechanical stimuli in dynamic conditions, and for this reason different techniques have been developed to both mechanically stimulate cells and record their response. The use of driven mechanical probes (e.g., AFM tips) (Charras and Horton 2002; Chaudhuri et al. 2009), patch pipettes (Hu and Lewin 2006; Kwan et al. 2009), magnetic
46
Biomed Microdevices (2015) 17:46
Page 2 of 11
tweezers (Glogauer et al. 1995; N Wang 1993; Sniadecki 2009), or optical tweezers (Cojoc et al. 2007; Resnick 2010) in association with patch clamp (Sakmann and Neher 1984) technique have been largely exploited. In such experiments, mechanotransduction processes at molecular level have been investigated, but also voltage (action potentials), and/or currents have been recorded with high precision, even recording from single channels (Delmas et al. 2011; Hamill et al. 1981; Sakmann and Neher 1984). However, such techniques are limited to single cells and they are very invasive. Moreover, only small deflections can be obtained (~1–2 μm) whereas in nature cell deformations can be much larger (e.g., muscle cells) (Kim et al. 2009). Other systems provide a way to stimulate large parts of cell membranes, involving also cell populations, by exploiting swelling phenomena or fluid shear stress (Haeberle et al. 2008; Pavalko et al. 1998; Tkachenko et al. 2009; Viana et al. 2001); or, cells can be stretched through the deformation of flexible substrates where they are cultured on (Brown 2000). This approach has been explored in several works (Bhattacharya et al. 2008; Heo et al. 2009; Mann et al. 2012; Wang et al. 2010; Chen et al. 2013) where the deformation was obtained by directly applying a stretch or a pressure on an elastomeric membrane. In these studies mechanotransduction was assessed from a biochemical and biomechanical point of view. The biochemical cell response to external mechanical stimuli was studied by the calcium imaging technique. Such method allows for the staining of calcium flux through the use of a sensitive dye in order to observe variations of Ca2+ concentrations into the cells during the stimulation. The biomechanical response is evaluated by the study of the re-organization of stress fibers, focal adhesions (FAs) and cytoskeleton (CSK) through imaging techniques. Indeed, during mechanotransduction an electrical signal is produced, and if recordings of its variations were performed, this could represent a viable method to evaluate cell response to mechanical stimulation in parallel and/or in correlation to biochemical and biomechanical analysis. Conductive substrates, electrodes, or electronic devices (e.g., organic based) have been widely used in cell studies, applied in very innovative systems for cell mechanical stimulation (Akbari and Shea 2012) or interfaced with cells for signal recording (Demelas et al. 2013; Heer et al. 2007; Hierlemann et al. 2011; Offenhäusser and Knoll 2001). In particular, one of the most known techniques is the electric cell-based impedance sensing (ECIS) (Giaever and Keese 1984). Here cells are cultured on rigid electrode surfaces and the electrode impedance is recorded, the latter varying according to cell attachment, movements, and morphology changes (Arndt et al. 2004; Giaever and Keese 1993). ECIS is a noninvasive method that allows assessment of cell kinetics, spreading and cytotoxicity (Arndt et al. 2004; Opp et al. 2009; Wegener et al. 2000). However, only recordings in static conditions can be addressed, due to the use of rigid metal
electrodes. Instead, the detection of cell response to mechanical stimuli should be pursued in dynamic conditions by recording the variation of the impedance of a deformable conductive substrate, caused by the variation of ionic current fluxes that occurs during the mechanotransduction process. In this work we present a soft, stretchable and conductive biointerface that represents the first step toward the fabrication of a new device for mechanotransduction studies, based on the combination of cell mechanical stimulation and of dynamic impedance recording. The biointerface is designed in such a way that cells can be cultured on its surface and mechanically stimulated by means of uniaxial strain, externally applied to the same biointerface. The latter is composed of two parts: i) a poly(dimethyl siloxane) (PDMS) soft and stretchable substrate; and ii) a conjugated polymer poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) conductive surface through which impedance recordings are performed to investigate cell response. Despite the lower stiffness compared to rigid metal electrodes, PEDOT:PSS lacks stretchability. For this reason, we exploited the self-assembling phenomenon of surface wrinkle formation (Chung et al. 2011; Fu et al. 2009; Genzer and Groenewold 2006) to allow for stretching the bilayer composite PDMS/PEDOT:PSS without cracking or loss of conductivity. Figure 1 depicts the biointerface concept together with a dedicated platform, integrating the biointerface, which was built and used to apply mechanical stimulation and record cellular response. In this investigation we report on the biointerface design and fabrication process. Secondly, we present the assessment of the morphological and electromechanical aspects. Then biocompatibility assessment and differentiation of SH-SY5Y human neuroblastoma cells is described. Finally, we present and discuss preliminary experiments on the recording of impedance variation occurring when the bio-hybrid construct (i.e., integrating the biointerface and the cells) is uniaxially strained.
2 Materials and methods 2.1 Materials Silicon wafers (400 μm thick, p type, boron doped, ) were purchased from Si-Mat Silicon Materials. Silicone elastomer poly(dimethyl siloxane) (PDMS, Sylgard 184) was purchased from Dow Corning Corp.. Conjugated polymer poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) dispersion in water (Clevios PH1000, PEDOT:PSS 1:2.5 w/w ratio) was purchased from Heraeus Gmbh, Germany, and was filtered through 1.20 μm pores size filters (Minisart®, Sartorius) before use. Chlorotrimethylsilane (99 %), dimethylsulfoxide (DMSO, ACS Reagent, 99.9 %),
Biomed Microdevices (2015) 17:46
Page 3 of 11 46
Fig. 1 Biointerface concept (a): the mechanical stimulation of cells is obtained by uniaxial stretching of the biointerface composed of a corrugated PEDOT:PSS surface [blue] and a soft PDMS substrate [gray]. Mechanical stimulation and impedance recording are performed through a dedicated platform, mainly equipped with a load cell and a translation stage, and an LCR meter (b). The platform can be used under a microscope to perform cell imaging studies (c)
(3-aminopropyl) triethoxysilane (APTES, 99 %) and ethanol (96 %) were purchased from Sigma Aldrich, Germany. Ammonia solution (30 %) was purchased from Carlo Erba, Italy. All chemicals were used as received. Silicon wafers were used for the fabrication of PDMS substrates; microscope glass slides were used as substrates for biointerfaces fabrication and treatments. De-ionized water was used in PDMS silanization treatment. 2.2 Fabrication of soft conductive stretchable biointerfaces The fabrication process of the biointerface is schematically reported in Fig. 2. Silicon wafers were silanized (Fig. 2a), to favor peeling off of the PDMS layers, by 30 min exposure to chlorotrimethylsilane vapor created in an airtight plastic container by putting a small quantity of solution in a vial. Silicone elastomer films were prepared by spin-coating PDMS (10:1 monomer/curing agent ratio) mixture on silicon wafers at 500 rpm for 60 s (Fig. 2b) and curing in a convection oven (Welland WF30) at T=90 °C for 1 h. The obtained PDMS films were cut with a razor blade (Fig. 2c), obtaining 2.0×
3.5 cm samples that were peeled off the wafer (Fig. 2d), transferred on a glass slide and pre-stretched to 15 % of their length by applying uniaxial strain along sample length direction (Fig. 2e). Sample surface was then chemically modified (Ciofani et al. 2012; Genchi et al. 2013) to render it hydrophilic and to prevent surface hydrophobic recovery: PDMS substrates were exposed to O2 plasma (Fig. 2f) for 60 s at 30 W, 0.5 mbar in a plasma system (Colibrì, Gambetti, Binasco, Italy), and incubated with a 60 % ethanol, 20 % (3aminopropyl) triethoxysilane (APTES), 1.2 % NH3 water solution (pH=9) for 3 h at 60 °C (Fig. 2g). This step was necessary in order to allow the PEDOT:PSS solution to spread and wet the surface in subsequent processing. Samples were then rinsed with ethanol and de-ionized water and dried with air flux. Dimethylsulfoxide (DMSO) 5 % w/w was added to the previously filtered PEDOT:PSS dispersion in water in order to increase its conductivity (Elschner 2011). After agitation for 8 h at room temperature with a magnetic stirrer, the doped PEDOT:PSS solution was filtered again and deposited on PDMS sample surface by two subsequent spin-coating (2500 rpm, 60 s) and drying (T=170 °C, 1 h) steps (Fig. 2h–j).
46
Page 4 of 11
Biomed Microdevices (2015) 17:46
Fig. 2 Scheme of the biointerface fabrication process: PDMS substrate preparation and surface chemical modification (a–g), surface conductive layer deposition (h–j), sample release and spontaneous formation of
wrinkles (k, l). In (l) the characteristic spatial periodicity of the wrinkled surface is indicated with λ
Finally, samples were cut in size of 2.0×1.5 cm (2 k) and released; micro-wrinkles (Fig. 2l) were formed on their surface, aligned along the direction perpendicular to the prestretching one.
2.4 SH-SY5Y cell culture
2.3 Biointerface surface morphology characterization The thickness (ts) of the PDMS substrate was measured with a Surfcom 130A profilometer (Carl Zeiss AG – Oberkochen, Germany) at the substrate edge. The thickness (tf) of the conductive film was derived from the characterization carried out in a previous work (Greco et al. 2013b) where it was measured by means of a P-6 stylus profilometer (KLA Tencor, USA). Surface topography of the biointerface was characterized by Atomic Force Microscope (AFM) imaging (Veeco Innova Scanning Probe Microscope). The images were collected operating in tapping mode, with oxide sharpened silicon probes (RTESPA-CP) at resonant frequency of ≈300 kHz. Average on plane spatial wrinkles periodicity (λ) and amplitude (A) were obtained by software analysis (Gwyddion SPM analysis tool) of the obtained AFM images. SEM images were obtained with a Helios NanoLab 600i Dual Beam FIB/FE-SEM (FEI, USA), operating at 5.0 kV accelerating voltage.
Human neuroblastoma SH-SY5Y cells (ATCC CRL-2266) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) and Ham’s F12 (1:1) with 10 % fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin and 2 mM Lglutamine. Cells were maintained at 37 °C in a saturated humidity atmosphere containing 95 % air / 5 % CO2. Differentiation of SH-SY5Y cells was induced incubating cells (10,000 /cm2) with a low-serum medium (DMEM with 1 % fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine) supplemented with 10 μM of all-trans retinoic acid. 2.5 Biological characterization 2.5.1 WST-1 and Live-Dead® assays For viability testing, cells were seeded on the substrates (5000 cells/cm2) positioned in 24-well plates and incubated for 3 and 6 days since seeding. Thereafter, cell metabolism was assessed with the WST-1 assay (2-(4-iodophenyl)-3-(4-nitophenyl)5-(2,4-disulfophenyl)-2H-tetrazoilium monosodium salt, provided in a pre-mix electro-coupling solution, BioVision). Cell cultures were treated with 300 μl of culture medium added with 30 μl of the pre-mix solution for 2 h and, finally,
Biomed Microdevices (2015) 17:46
absorbance was read at 450 nm with a microplate reader (Victor3, Perkin Elmer). Viability was further qualitatively investigated at the same time-points with the Live/Dead® viability/cytotoxicity Kit (Molecular Probes). The kit contains calcein AM (4 mM in anhydrous DMSO) and ethidium homodimer-1 (EthD-1, 2 mM in DMSO/H2O 1:4v/v), and allows for the discrimination between live cells (stained in green by calcein) and dead cells (stained in red by EthD-1). Cultures were rinsed with PBS, treated for 10 min at 37 °C with 2 μM calcein AM and 4 μM EthD-1 in PBS, and finally observed with an inverted fluorescence microscope (TE2000U, Nikon) equipped with a cooled CCD camera (DS-5MC USB2, Nikon) and with NIS Elements imaging software. Appropriate control cultures on standard surfaces (polystyrene dishes) were carried out. 2.5.2 Immunofluorescence After 5 days of differentiation, cultures were tested for β3tubulin expression (a typical neuronal marker). Samples were rinsed with PBS and fixed in paraformaldehyde (4 % in PBS) for 15 min. After rinsing with PBS, they were incubated with sodium borohydryde (1 mg/ml in PBS) for 10 min to reduce autofluorescence and cell membranes were thereafter permeabilized with 0.1 % Triton X-100 in PBS for 15 min. Antibody aspecific binding sites were saturated with 10 % goat serum in PBS for 1 h, and, subsequently, the primary antibody (rabbit polyclonal IgG anti-tubulin, Sigma, diluted 1:75 in 10 % goat serum) was added. After 30 min of incubation at 37 °C, samples were rinsed with 10 % goat serum; then, a staining solution was added, composed of a secondary antibody (fluorescent goat anti-rabbit IgG, Invitrogen) diluted 1:250 in 10 % goat serum, of 100 μM of TRITC-phalloidin (Sigma) for factin staining, and of 1 μM DAPI for nucleus counterstaining. After 30 min of incubation at room temperature, samples were rinsed with 0.45 M NaCl in PBS for 1 min to remove weakly bound antibodies and, after rinsing in PBS, observed with the inverted fluorescence microscope. 2.6 Electromechanical characterization and impedance variation recording The electrical behavior of the biointerface under the effect of a mechanical stimulation was studied by means of a custom platform, in which the biointerface was integrated, depicted in Fig. 1b, c. Uniaxial strain was applied to the biointerface by means of a micrometric slider (M-126.CG1, PI, USA), in a cyclic manner, along the axis perpendicular to the direction of surface wrinkles. All parameters (i.e., velocity, acceleration, displacement, and durations at both final stretched state and initial state) were set and controlled via a software (developed in VisualStudio2008) and a motor controller (c-863
Page 5 of 11 46
Mercury™ servo motor controller, PI, USA). Opposite to the slider, a load cell (LRF400, Futek, USA) allowed to simultaneously record the force required for the uniaxial deformation. The biointerface was linked to the slider and the load cell by means of two s-shaped arms holding the two ends of the biointerface immersed in the culture medium. The electrical impedance at frequency 10 KHz was simultaneously recorded by means of two metal electrodes connected to an LCR meter (E4948A, Agilent Technologies, USA) and clamped to sample edges. All data were recorded by means of an acquisition system (NI USB-6218, National Instruments, USA). Once attached to the slider and the load cell, samples were stretched to a 10 % maximum strain with a 0.6 mm/s velocity, that is, samples reached the maximum strain in a displacement time of 1.67 s. The time parameter that was set for each half of the cycle was of 25 s, which included the displacement time and the time at final state (either stretched or initial). The mechanical stimulation of cells was performed by deforming the biointerface on which they were cultured on. Stretchability of the substrates allowed transmission of the stimulus to the cells, which were cultured on a surface of 1 cm2 at the center of the sample. Cell differentiation was induced after 1 day of culture, and the stretching experiment was performed after 5 days of differentiation. During the experiment, the culture medium was changed with a CO2-independent medium in order to maintain cells viability out of the incubator. All culture treatments and experimental procedures were performed in parallel on control samples without cells cultured on the surface. Impedance was continuously recorded during the experiment, and the relative impedance variation ΔZr = (Z s -Z i )/Z i was calculated for each trial (Z s is the biointerface impedance in the final stretched state, Zi in the initial state). 2.7 Statistical analysis Biological data were studied with analysis of variance (ANOVA) followed by Bonferroni’s post-hoc test to evaluate significance, that was set at p < 0.05, with the aid of KaleidaGraph (Sinergy Software). In all cases, three independent experiments were carried out. Results are presented as mean value±standard deviation. Moreover, data resulting from impedance recording were analyzed with a t-test.
3 Results and discussion 3.1 Biointerfaces fabrication and wrinkled surface morphology characterization Stretchable and conductive biointerfaces were obtained, as previously described (Bernardeschi et al. 2013), by the deposition of a thin film of PEDOT:PSS doped with 5 % DMSO on
46
Page 6 of 11
an elastic PDMS substrate. The choice of these materials was driven by the need to have a flexible, deformable, soft and conductive system. However, the materials also had to fulfill other requirements, like biocompatibility, as well as optical transparency needed for microscope imaging. In particular, PDMS was selected as the soft and stretchable material for the substrate because of its known properties (Sia and Whitesides 2003). Moreover, it is easy to process because it can be deposited by spin-coating and its mechanical properties can be simply modulated by varying the monomer/curing agent ratio. For the conductive surface, the conjugated polymer PEDOT:PSS was chosen because it presents a good electrical conductivity that can be improved by means of doping agents, is optically transparent, and can be deposited by spincoating from an aqueous solution. Moreover, the biocompatibility and the functional behavior as a cell-contact electrode of PEDOT:PSS has been tested in many different configurations and with several kinds of cells (Berggren and RichterDahlfors 2007; Blau et al. 2011; Greco et al. 2013a; Guimard et al. 2007). Given the hydrophobicity of PDMS, it was necessary to activate the PDMS surface: oxygen plasma was first used to both make the PDMS surface wettable and to obtain a uniform conductive layer, and this method was successfully verified. Nonetheless, during the culturing phase at specific conditions (i.e., continuous, long term immersion in liquid cell culture medium), PEDOT:PSS films partially detached from the PDMS surface and the wrinkles disappeared. This was maybe due to the hydrophobic recovery of PDMS and to a poor adhesion between layers affected by the culture medium. Deposition of PEDOT:PSS on hydrophobic substrates has been previously adopted in the fabrication of stretchable electronic devices (e.g., solar cells) (Lipomi et al. 2011, 2012; Oh et al. 2014; Vosgueritchian et al. 2012). Authors of the cited works reported that it was necessary to add a fluorosurfactant to the specific PEDOT:PSS formulation, in order to allow good wetting by the waterborne PEDOT:PSS solution and to obtain a uniform conductive layer. For this reason, before performing the spin coating of the PEDOT:PSS layer, the PDMS surface was irreversibly chemically modified by means of a silanization treatment with a water solution of an alkoxysilane, APTES (Ciofani et al. 2012; Genchi et al. 2013). Following this, the so obtained samples could maintain their integrity and showed stable surface patterns after 6 days in the given culture conditions. The latter is a sufficient time span requested for cell viability and proliferation assessments, and to obtain differentiated cells. In this work, it was fundamental to exploit surface wrinkling (Chung et al. 2011; Fu et al. 2009; Genzer and Groenewold 2006) in order for the non-stretchable PEDOT:PSS layer to sustain the applied stimulus without loss of conductivity. In this way, micro-corrugations aligned along a specific direction, and having a tunable spatial periodicity,
Biomed Microdevices (2015) 17:46
were formed on the surface during fabrication; wrinkles extended during the applied stretching, thus avoiding the cracking of the conducting surface. In order to induce formation of micro-wrinkles, the PEDOT:PSS solution was deposited by spin-coating on a stretched PDMS substrate (15 % prestretching amount), and dried to obtain a uniform and smooth conducting polymer Bskin^. The compressive forces acting on this skin, during the release from the pre-stretched state, induced the formation of micro-wrinkles on the surface that were aligned perpendicularly to the applied strain. In fact, when a stiff film is deposited on a soft stretchable substrate and compressed, the bending energy of the compressed rigid film competes with the deformation energy of the substrate, leading to wrinkles formation on the surface at the equilibrium (Chung et al. 2011; Genzer and Groenewold 2006). The biointerface surface was characterized by SEM imaging, as shown in Fig. 3. In Fig. 3b, a corrugated conductive film of PEDOT:PSS (thickness tf =90 nm) is visible above the nonconductive PDMS substrate (thickness ts =130 μm). A qualitative evaluation also showed the presence of cracks, in some points of the biointerface surface, parallel to the pre-stretch direction. Despite this, PEDOT:PSS films showed electrical continuity. A quantitative estimation of the surface topography was obtained by the analysis of the acquired AFM images (Fig. 4a). The on-plane spatial periodicity λw of wrinkles was derived from the horizontal Power Spectral Density Function (PSDF) calculated on the surface topography signal. It represents the power distribution of such signal in the reciprocal space k that is inversely proportional to the spatial periodicity (k=2π/λ). In Fig. 4b, the normalized distribution is reported as a function of λ: the value of λ in correspondence of the peak of distribution represents the value of maximum correlation of the signal, corresponding to the representative wavelength λw. The obtained wrinkle spatial periodicity is λw =4.37±0.17 μm. Wrinkle amplitude was calculated as the average absolute value Rz of the five highest peaks and the five lowest valleys, over the horizontal profiles perpendicular to wrinkles direction, at different positions of the biointerface surface. An amplitude value of A=0.598±0.014 μm was obtained. The experimental findings can be compared with the theoretical values of the wrinkle amplitude and wavelength (Chung et al. 2011). A theoretical value of λ was calculated by using the following values for Young’s modulus and Poisson’s ratio, for substrate and surface film: Es =2 MPa, νs =0.5, Ef =2.3 GPa and νf =0.35 (Tahk et al. 2009). The obtained value is λ=4.32 μm, that is in reasonable accordance with the experimental findings. The calculated critical strain needed to have wrinkles formation is εc =0.005. Considering the applied pre-strain value of ε=0.15, the obtained theoretical amplitude value is A= 0.538 μm. Also in this case, accordance can be found between theoretical and experimental data.
Biomed Microdevices (2015) 17:46
Page 7 of 11 46
Fig. 3 Uniaxially-aligned wrinkles on the PDMS/PEDOT:PSS biointerface surface depicted by SEM analysis (a); the inset shows the fast Fourier transform of the image. Cross-sectional view in correspondence to a vertical cut along the stretching direction (b)
The results of surface morphology characterization allowed the process to be validated as capable of producing stable and uniform micro-wrinkles all over the samples surface, with characteristic topographical features (wrinkles on plane spatial periodicity and amplitude) in a very reproducible way. Moreover, the proposed process permits an easy tailoring of such features in order to meet different requirements in future applications, by a clever design of layers’ thicknesses, Young’s moduli ratio of film and substrate and pre-stretching amount. 3.2 Biocompatibility and cellular differentiation As shown in Fig. 5, biointerfaces well supported cell viability, which can be appreciated by the green fluorescent staining in Fig. 5a, with less than 2 % dead cells, stained in red. Moreover, WST-1 assay highlighted excellent cell metabolic activity on the biointerfaces (Fig. 5b), that, after 6 days of proliferation, became higher of about 25 % (p
A soft, stretchable and conductive biointerface for cell mechanobiology Irene Bernardeschi & Francesco Greco & Gianni Ciofani & Attilio Marino & Virgilio Mattoli & Barbara Mazzolai & Lucia Beccai
# Springer Science+Business Media New York 2015
Abstract In mechanobiology the study of cell response to mechanical stimuli is fundamental, and the involved processes (i.e., mechanotransduction) need to be investigated by interfacing (mechanically and electrically) with the cells in dynamic and non-invasive natural-like conditions. In this work, we present a novel soft, stretchable and conductive biointerface that allows both cell mechanical stimulation and dynamic impedance recording. The biointerface stretchability and conductivity, jointly to the biocompatibility and transparency needed to perform cell culture studies, were obtained by exploiting the formation of wrinkles on the surface of a 90 nm thick conductive layer of poly(3,4ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) on a pre-stretched 130 μm thick poly(dimethylsiloxane) (PDMS) substrate. Cell adhesion and proliferation of SH-SY5Y human neuroblastoma cells were evaluated, and cell differentiation on the corrugated surface was assessed. We demonstrate how the biointerface remains conductive when applying uniaxial strain up to 10 %, and when cell culturing is performed. Finally, a reduction of about 30 % of the relative impedance variation signal was measured, with respect to the control, as a result of the mechanical stimulation of cells.
Electronic supplementary material The online version of this article (doi:10.1007/s10544-015-9950-0) contains supplementary material, which is available to authorized users. I. Bernardeschi : F. Greco : G. Ciofani : A. Marino : V. Mattoli : B. Mazzolai : L. Beccai (*) Center for Micro-BioRobotics, Istituto Italiano di Tecnologia, Viale Rinaldo Piaggio 34, 56025 Pontedera, PI, Italy e-mail: [email protected] I. Bernardeschi : A. Marino The Biorobotics Institute, Scuola Superiore Sant’Anna, Viale Rinaldo Piaggio 34, 56025 Pontedera, Italy
Keywords Stretchable biointerface . Impedance . Biohybrid . Surface wrinkling . PEDOT:PSS . Mechanotransduction
1 Introduction Living organisms, from plants and microorganisms to mammals, are mechanosensitive that is they can feel and react to external forces in order to interact with the environment where they live. Mechanobiology studies the effect of such forces on cells and, in particular, is focused on the understanding of mechanotransduction, a basic process by which cells can convert external mechanical stimuli into biochemical signals. Mechanical forces regulate cells biochemistry and the correct functioning of physiological processes and homeostasis are maintained through the mechanotransduction process (Ingber 2003; Jaalouk and Lammerding 2009; Orr et al. 2006). Indeed, mechanotransduction studies have a significant role in recognizing causes for some pathologies associated to an incorrect regulation of physiological processes (e.g., a defect in protein expression due to an wrong transduction process) (Coughlin et al. 2008; Heydemann and McNally 2007). However, it is mostly in sensing that mechanotransduction plays a main role; specifically, in hearing and touch, where external mechanical forces - produced by, e.g., sound waves, pressure or stretch - act on the dedicated sensing organs (i.e., outer and inner ear and skin), where they are detected and transformed into electrochemical signals. Investigating mechanotransduction requires the study of cell response to mechanical stimuli in dynamic conditions, and for this reason different techniques have been developed to both mechanically stimulate cells and record their response. The use of driven mechanical probes (e.g., AFM tips) (Charras and Horton 2002; Chaudhuri et al. 2009), patch pipettes (Hu and Lewin 2006; Kwan et al. 2009), magnetic
46
Biomed Microdevices (2015) 17:46
Page 2 of 11
tweezers (Glogauer et al. 1995; N Wang 1993; Sniadecki 2009), or optical tweezers (Cojoc et al. 2007; Resnick 2010) in association with patch clamp (Sakmann and Neher 1984) technique have been largely exploited. In such experiments, mechanotransduction processes at molecular level have been investigated, but also voltage (action potentials), and/or currents have been recorded with high precision, even recording from single channels (Delmas et al. 2011; Hamill et al. 1981; Sakmann and Neher 1984). However, such techniques are limited to single cells and they are very invasive. Moreover, only small deflections can be obtained (~1–2 μm) whereas in nature cell deformations can be much larger (e.g., muscle cells) (Kim et al. 2009). Other systems provide a way to stimulate large parts of cell membranes, involving also cell populations, by exploiting swelling phenomena or fluid shear stress (Haeberle et al. 2008; Pavalko et al. 1998; Tkachenko et al. 2009; Viana et al. 2001); or, cells can be stretched through the deformation of flexible substrates where they are cultured on (Brown 2000). This approach has been explored in several works (Bhattacharya et al. 2008; Heo et al. 2009; Mann et al. 2012; Wang et al. 2010; Chen et al. 2013) where the deformation was obtained by directly applying a stretch or a pressure on an elastomeric membrane. In these studies mechanotransduction was assessed from a biochemical and biomechanical point of view. The biochemical cell response to external mechanical stimuli was studied by the calcium imaging technique. Such method allows for the staining of calcium flux through the use of a sensitive dye in order to observe variations of Ca2+ concentrations into the cells during the stimulation. The biomechanical response is evaluated by the study of the re-organization of stress fibers, focal adhesions (FAs) and cytoskeleton (CSK) through imaging techniques. Indeed, during mechanotransduction an electrical signal is produced, and if recordings of its variations were performed, this could represent a viable method to evaluate cell response to mechanical stimulation in parallel and/or in correlation to biochemical and biomechanical analysis. Conductive substrates, electrodes, or electronic devices (e.g., organic based) have been widely used in cell studies, applied in very innovative systems for cell mechanical stimulation (Akbari and Shea 2012) or interfaced with cells for signal recording (Demelas et al. 2013; Heer et al. 2007; Hierlemann et al. 2011; Offenhäusser and Knoll 2001). In particular, one of the most known techniques is the electric cell-based impedance sensing (ECIS) (Giaever and Keese 1984). Here cells are cultured on rigid electrode surfaces and the electrode impedance is recorded, the latter varying according to cell attachment, movements, and morphology changes (Arndt et al. 2004; Giaever and Keese 1993). ECIS is a noninvasive method that allows assessment of cell kinetics, spreading and cytotoxicity (Arndt et al. 2004; Opp et al. 2009; Wegener et al. 2000). However, only recordings in static conditions can be addressed, due to the use of rigid metal
electrodes. Instead, the detection of cell response to mechanical stimuli should be pursued in dynamic conditions by recording the variation of the impedance of a deformable conductive substrate, caused by the variation of ionic current fluxes that occurs during the mechanotransduction process. In this work we present a soft, stretchable and conductive biointerface that represents the first step toward the fabrication of a new device for mechanotransduction studies, based on the combination of cell mechanical stimulation and of dynamic impedance recording. The biointerface is designed in such a way that cells can be cultured on its surface and mechanically stimulated by means of uniaxial strain, externally applied to the same biointerface. The latter is composed of two parts: i) a poly(dimethyl siloxane) (PDMS) soft and stretchable substrate; and ii) a conjugated polymer poly(3,4ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS) conductive surface through which impedance recordings are performed to investigate cell response. Despite the lower stiffness compared to rigid metal electrodes, PEDOT:PSS lacks stretchability. For this reason, we exploited the self-assembling phenomenon of surface wrinkle formation (Chung et al. 2011; Fu et al. 2009; Genzer and Groenewold 2006) to allow for stretching the bilayer composite PDMS/PEDOT:PSS without cracking or loss of conductivity. Figure 1 depicts the biointerface concept together with a dedicated platform, integrating the biointerface, which was built and used to apply mechanical stimulation and record cellular response. In this investigation we report on the biointerface design and fabrication process. Secondly, we present the assessment of the morphological and electromechanical aspects. Then biocompatibility assessment and differentiation of SH-SY5Y human neuroblastoma cells is described. Finally, we present and discuss preliminary experiments on the recording of impedance variation occurring when the bio-hybrid construct (i.e., integrating the biointerface and the cells) is uniaxially strained.
2 Materials and methods 2.1 Materials Silicon wafers (400 μm thick, p type, boron doped, ) were purchased from Si-Mat Silicon Materials. Silicone elastomer poly(dimethyl siloxane) (PDMS, Sylgard 184) was purchased from Dow Corning Corp.. Conjugated polymer poly(3,4-ethylenedioxythiophene):poly(styrene sulfonate) (PEDOT:PSS) dispersion in water (Clevios PH1000, PEDOT:PSS 1:2.5 w/w ratio) was purchased from Heraeus Gmbh, Germany, and was filtered through 1.20 μm pores size filters (Minisart®, Sartorius) before use. Chlorotrimethylsilane (99 %), dimethylsulfoxide (DMSO, ACS Reagent, 99.9 %),
Biomed Microdevices (2015) 17:46
Page 3 of 11 46
Fig. 1 Biointerface concept (a): the mechanical stimulation of cells is obtained by uniaxial stretching of the biointerface composed of a corrugated PEDOT:PSS surface [blue] and a soft PDMS substrate [gray]. Mechanical stimulation and impedance recording are performed through a dedicated platform, mainly equipped with a load cell and a translation stage, and an LCR meter (b). The platform can be used under a microscope to perform cell imaging studies (c)
(3-aminopropyl) triethoxysilane (APTES, 99 %) and ethanol (96 %) were purchased from Sigma Aldrich, Germany. Ammonia solution (30 %) was purchased from Carlo Erba, Italy. All chemicals were used as received. Silicon wafers were used for the fabrication of PDMS substrates; microscope glass slides were used as substrates for biointerfaces fabrication and treatments. De-ionized water was used in PDMS silanization treatment. 2.2 Fabrication of soft conductive stretchable biointerfaces The fabrication process of the biointerface is schematically reported in Fig. 2. Silicon wafers were silanized (Fig. 2a), to favor peeling off of the PDMS layers, by 30 min exposure to chlorotrimethylsilane vapor created in an airtight plastic container by putting a small quantity of solution in a vial. Silicone elastomer films were prepared by spin-coating PDMS (10:1 monomer/curing agent ratio) mixture on silicon wafers at 500 rpm for 60 s (Fig. 2b) and curing in a convection oven (Welland WF30) at T=90 °C for 1 h. The obtained PDMS films were cut with a razor blade (Fig. 2c), obtaining 2.0×
3.5 cm samples that were peeled off the wafer (Fig. 2d), transferred on a glass slide and pre-stretched to 15 % of their length by applying uniaxial strain along sample length direction (Fig. 2e). Sample surface was then chemically modified (Ciofani et al. 2012; Genchi et al. 2013) to render it hydrophilic and to prevent surface hydrophobic recovery: PDMS substrates were exposed to O2 plasma (Fig. 2f) for 60 s at 30 W, 0.5 mbar in a plasma system (Colibrì, Gambetti, Binasco, Italy), and incubated with a 60 % ethanol, 20 % (3aminopropyl) triethoxysilane (APTES), 1.2 % NH3 water solution (pH=9) for 3 h at 60 °C (Fig. 2g). This step was necessary in order to allow the PEDOT:PSS solution to spread and wet the surface in subsequent processing. Samples were then rinsed with ethanol and de-ionized water and dried with air flux. Dimethylsulfoxide (DMSO) 5 % w/w was added to the previously filtered PEDOT:PSS dispersion in water in order to increase its conductivity (Elschner 2011). After agitation for 8 h at room temperature with a magnetic stirrer, the doped PEDOT:PSS solution was filtered again and deposited on PDMS sample surface by two subsequent spin-coating (2500 rpm, 60 s) and drying (T=170 °C, 1 h) steps (Fig. 2h–j).
46
Page 4 of 11
Biomed Microdevices (2015) 17:46
Fig. 2 Scheme of the biointerface fabrication process: PDMS substrate preparation and surface chemical modification (a–g), surface conductive layer deposition (h–j), sample release and spontaneous formation of
wrinkles (k, l). In (l) the characteristic spatial periodicity of the wrinkled surface is indicated with λ
Finally, samples were cut in size of 2.0×1.5 cm (2 k) and released; micro-wrinkles (Fig. 2l) were formed on their surface, aligned along the direction perpendicular to the prestretching one.
2.4 SH-SY5Y cell culture
2.3 Biointerface surface morphology characterization The thickness (ts) of the PDMS substrate was measured with a Surfcom 130A profilometer (Carl Zeiss AG – Oberkochen, Germany) at the substrate edge. The thickness (tf) of the conductive film was derived from the characterization carried out in a previous work (Greco et al. 2013b) where it was measured by means of a P-6 stylus profilometer (KLA Tencor, USA). Surface topography of the biointerface was characterized by Atomic Force Microscope (AFM) imaging (Veeco Innova Scanning Probe Microscope). The images were collected operating in tapping mode, with oxide sharpened silicon probes (RTESPA-CP) at resonant frequency of ≈300 kHz. Average on plane spatial wrinkles periodicity (λ) and amplitude (A) were obtained by software analysis (Gwyddion SPM analysis tool) of the obtained AFM images. SEM images were obtained with a Helios NanoLab 600i Dual Beam FIB/FE-SEM (FEI, USA), operating at 5.0 kV accelerating voltage.
Human neuroblastoma SH-SY5Y cells (ATCC CRL-2266) were cultured in Dulbecco’s modified Eagle’s medium (DMEM) and Ham’s F12 (1:1) with 10 % fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin and 2 mM Lglutamine. Cells were maintained at 37 °C in a saturated humidity atmosphere containing 95 % air / 5 % CO2. Differentiation of SH-SY5Y cells was induced incubating cells (10,000 /cm2) with a low-serum medium (DMEM with 1 % fetal bovine serum, 100 IU/ml penicillin, 100 μg/ml streptomycin, and 2 mM L-glutamine) supplemented with 10 μM of all-trans retinoic acid. 2.5 Biological characterization 2.5.1 WST-1 and Live-Dead® assays For viability testing, cells were seeded on the substrates (5000 cells/cm2) positioned in 24-well plates and incubated for 3 and 6 days since seeding. Thereafter, cell metabolism was assessed with the WST-1 assay (2-(4-iodophenyl)-3-(4-nitophenyl)5-(2,4-disulfophenyl)-2H-tetrazoilium monosodium salt, provided in a pre-mix electro-coupling solution, BioVision). Cell cultures were treated with 300 μl of culture medium added with 30 μl of the pre-mix solution for 2 h and, finally,
Biomed Microdevices (2015) 17:46
absorbance was read at 450 nm with a microplate reader (Victor3, Perkin Elmer). Viability was further qualitatively investigated at the same time-points with the Live/Dead® viability/cytotoxicity Kit (Molecular Probes). The kit contains calcein AM (4 mM in anhydrous DMSO) and ethidium homodimer-1 (EthD-1, 2 mM in DMSO/H2O 1:4v/v), and allows for the discrimination between live cells (stained in green by calcein) and dead cells (stained in red by EthD-1). Cultures were rinsed with PBS, treated for 10 min at 37 °C with 2 μM calcein AM and 4 μM EthD-1 in PBS, and finally observed with an inverted fluorescence microscope (TE2000U, Nikon) equipped with a cooled CCD camera (DS-5MC USB2, Nikon) and with NIS Elements imaging software. Appropriate control cultures on standard surfaces (polystyrene dishes) were carried out. 2.5.2 Immunofluorescence After 5 days of differentiation, cultures were tested for β3tubulin expression (a typical neuronal marker). Samples were rinsed with PBS and fixed in paraformaldehyde (4 % in PBS) for 15 min. After rinsing with PBS, they were incubated with sodium borohydryde (1 mg/ml in PBS) for 10 min to reduce autofluorescence and cell membranes were thereafter permeabilized with 0.1 % Triton X-100 in PBS for 15 min. Antibody aspecific binding sites were saturated with 10 % goat serum in PBS for 1 h, and, subsequently, the primary antibody (rabbit polyclonal IgG anti-tubulin, Sigma, diluted 1:75 in 10 % goat serum) was added. After 30 min of incubation at 37 °C, samples were rinsed with 10 % goat serum; then, a staining solution was added, composed of a secondary antibody (fluorescent goat anti-rabbit IgG, Invitrogen) diluted 1:250 in 10 % goat serum, of 100 μM of TRITC-phalloidin (Sigma) for factin staining, and of 1 μM DAPI for nucleus counterstaining. After 30 min of incubation at room temperature, samples were rinsed with 0.45 M NaCl in PBS for 1 min to remove weakly bound antibodies and, after rinsing in PBS, observed with the inverted fluorescence microscope. 2.6 Electromechanical characterization and impedance variation recording The electrical behavior of the biointerface under the effect of a mechanical stimulation was studied by means of a custom platform, in which the biointerface was integrated, depicted in Fig. 1b, c. Uniaxial strain was applied to the biointerface by means of a micrometric slider (M-126.CG1, PI, USA), in a cyclic manner, along the axis perpendicular to the direction of surface wrinkles. All parameters (i.e., velocity, acceleration, displacement, and durations at both final stretched state and initial state) were set and controlled via a software (developed in VisualStudio2008) and a motor controller (c-863
Page 5 of 11 46
Mercury™ servo motor controller, PI, USA). Opposite to the slider, a load cell (LRF400, Futek, USA) allowed to simultaneously record the force required for the uniaxial deformation. The biointerface was linked to the slider and the load cell by means of two s-shaped arms holding the two ends of the biointerface immersed in the culture medium. The electrical impedance at frequency 10 KHz was simultaneously recorded by means of two metal electrodes connected to an LCR meter (E4948A, Agilent Technologies, USA) and clamped to sample edges. All data were recorded by means of an acquisition system (NI USB-6218, National Instruments, USA). Once attached to the slider and the load cell, samples were stretched to a 10 % maximum strain with a 0.6 mm/s velocity, that is, samples reached the maximum strain in a displacement time of 1.67 s. The time parameter that was set for each half of the cycle was of 25 s, which included the displacement time and the time at final state (either stretched or initial). The mechanical stimulation of cells was performed by deforming the biointerface on which they were cultured on. Stretchability of the substrates allowed transmission of the stimulus to the cells, which were cultured on a surface of 1 cm2 at the center of the sample. Cell differentiation was induced after 1 day of culture, and the stretching experiment was performed after 5 days of differentiation. During the experiment, the culture medium was changed with a CO2-independent medium in order to maintain cells viability out of the incubator. All culture treatments and experimental procedures were performed in parallel on control samples without cells cultured on the surface. Impedance was continuously recorded during the experiment, and the relative impedance variation ΔZr = (Z s -Z i )/Z i was calculated for each trial (Z s is the biointerface impedance in the final stretched state, Zi in the initial state). 2.7 Statistical analysis Biological data were studied with analysis of variance (ANOVA) followed by Bonferroni’s post-hoc test to evaluate significance, that was set at p < 0.05, with the aid of KaleidaGraph (Sinergy Software). In all cases, three independent experiments were carried out. Results are presented as mean value±standard deviation. Moreover, data resulting from impedance recording were analyzed with a t-test.
3 Results and discussion 3.1 Biointerfaces fabrication and wrinkled surface morphology characterization Stretchable and conductive biointerfaces were obtained, as previously described (Bernardeschi et al. 2013), by the deposition of a thin film of PEDOT:PSS doped with 5 % DMSO on
46
Page 6 of 11
an elastic PDMS substrate. The choice of these materials was driven by the need to have a flexible, deformable, soft and conductive system. However, the materials also had to fulfill other requirements, like biocompatibility, as well as optical transparency needed for microscope imaging. In particular, PDMS was selected as the soft and stretchable material for the substrate because of its known properties (Sia and Whitesides 2003). Moreover, it is easy to process because it can be deposited by spin-coating and its mechanical properties can be simply modulated by varying the monomer/curing agent ratio. For the conductive surface, the conjugated polymer PEDOT:PSS was chosen because it presents a good electrical conductivity that can be improved by means of doping agents, is optically transparent, and can be deposited by spincoating from an aqueous solution. Moreover, the biocompatibility and the functional behavior as a cell-contact electrode of PEDOT:PSS has been tested in many different configurations and with several kinds of cells (Berggren and RichterDahlfors 2007; Blau et al. 2011; Greco et al. 2013a; Guimard et al. 2007). Given the hydrophobicity of PDMS, it was necessary to activate the PDMS surface: oxygen plasma was first used to both make the PDMS surface wettable and to obtain a uniform conductive layer, and this method was successfully verified. Nonetheless, during the culturing phase at specific conditions (i.e., continuous, long term immersion in liquid cell culture medium), PEDOT:PSS films partially detached from the PDMS surface and the wrinkles disappeared. This was maybe due to the hydrophobic recovery of PDMS and to a poor adhesion between layers affected by the culture medium. Deposition of PEDOT:PSS on hydrophobic substrates has been previously adopted in the fabrication of stretchable electronic devices (e.g., solar cells) (Lipomi et al. 2011, 2012; Oh et al. 2014; Vosgueritchian et al. 2012). Authors of the cited works reported that it was necessary to add a fluorosurfactant to the specific PEDOT:PSS formulation, in order to allow good wetting by the waterborne PEDOT:PSS solution and to obtain a uniform conductive layer. For this reason, before performing the spin coating of the PEDOT:PSS layer, the PDMS surface was irreversibly chemically modified by means of a silanization treatment with a water solution of an alkoxysilane, APTES (Ciofani et al. 2012; Genchi et al. 2013). Following this, the so obtained samples could maintain their integrity and showed stable surface patterns after 6 days in the given culture conditions. The latter is a sufficient time span requested for cell viability and proliferation assessments, and to obtain differentiated cells. In this work, it was fundamental to exploit surface wrinkling (Chung et al. 2011; Fu et al. 2009; Genzer and Groenewold 2006) in order for the non-stretchable PEDOT:PSS layer to sustain the applied stimulus without loss of conductivity. In this way, micro-corrugations aligned along a specific direction, and having a tunable spatial periodicity,
Biomed Microdevices (2015) 17:46
were formed on the surface during fabrication; wrinkles extended during the applied stretching, thus avoiding the cracking of the conducting surface. In order to induce formation of micro-wrinkles, the PEDOT:PSS solution was deposited by spin-coating on a stretched PDMS substrate (15 % prestretching amount), and dried to obtain a uniform and smooth conducting polymer Bskin^. The compressive forces acting on this skin, during the release from the pre-stretched state, induced the formation of micro-wrinkles on the surface that were aligned perpendicularly to the applied strain. In fact, when a stiff film is deposited on a soft stretchable substrate and compressed, the bending energy of the compressed rigid film competes with the deformation energy of the substrate, leading to wrinkles formation on the surface at the equilibrium (Chung et al. 2011; Genzer and Groenewold 2006). The biointerface surface was characterized by SEM imaging, as shown in Fig. 3. In Fig. 3b, a corrugated conductive film of PEDOT:PSS (thickness tf =90 nm) is visible above the nonconductive PDMS substrate (thickness ts =130 μm). A qualitative evaluation also showed the presence of cracks, in some points of the biointerface surface, parallel to the pre-stretch direction. Despite this, PEDOT:PSS films showed electrical continuity. A quantitative estimation of the surface topography was obtained by the analysis of the acquired AFM images (Fig. 4a). The on-plane spatial periodicity λw of wrinkles was derived from the horizontal Power Spectral Density Function (PSDF) calculated on the surface topography signal. It represents the power distribution of such signal in the reciprocal space k that is inversely proportional to the spatial periodicity (k=2π/λ). In Fig. 4b, the normalized distribution is reported as a function of λ: the value of λ in correspondence of the peak of distribution represents the value of maximum correlation of the signal, corresponding to the representative wavelength λw. The obtained wrinkle spatial periodicity is λw =4.37±0.17 μm. Wrinkle amplitude was calculated as the average absolute value Rz of the five highest peaks and the five lowest valleys, over the horizontal profiles perpendicular to wrinkles direction, at different positions of the biointerface surface. An amplitude value of A=0.598±0.014 μm was obtained. The experimental findings can be compared with the theoretical values of the wrinkle amplitude and wavelength (Chung et al. 2011). A theoretical value of λ was calculated by using the following values for Young’s modulus and Poisson’s ratio, for substrate and surface film: Es =2 MPa, νs =0.5, Ef =2.3 GPa and νf =0.35 (Tahk et al. 2009). The obtained value is λ=4.32 μm, that is in reasonable accordance with the experimental findings. The calculated critical strain needed to have wrinkles formation is εc =0.005. Considering the applied pre-strain value of ε=0.15, the obtained theoretical amplitude value is A= 0.538 μm. Also in this case, accordance can be found between theoretical and experimental data.
Biomed Microdevices (2015) 17:46
Page 7 of 11 46
Fig. 3 Uniaxially-aligned wrinkles on the PDMS/PEDOT:PSS biointerface surface depicted by SEM analysis (a); the inset shows the fast Fourier transform of the image. Cross-sectional view in correspondence to a vertical cut along the stretching direction (b)
The results of surface morphology characterization allowed the process to be validated as capable of producing stable and uniform micro-wrinkles all over the samples surface, with characteristic topographical features (wrinkles on plane spatial periodicity and amplitude) in a very reproducible way. Moreover, the proposed process permits an easy tailoring of such features in order to meet different requirements in future applications, by a clever design of layers’ thicknesses, Young’s moduli ratio of film and substrate and pre-stretching amount. 3.2 Biocompatibility and cellular differentiation As shown in Fig. 5, biointerfaces well supported cell viability, which can be appreciated by the green fluorescent staining in Fig. 5a, with less than 2 % dead cells, stained in red. Moreover, WST-1 assay highlighted excellent cell metabolic activity on the biointerfaces (Fig. 5b), that, after 6 days of proliferation, became higher of about 25 % (p
