Research Article www.acsami.org
Enhanced In Vitro Biocompatibility of Chemically Modified Poly(dimethylsiloxane) Surfaces for Stable Adhesion and Long-term Investigation of Brain Cerebral Cortex Cells Shreyas Kuddannaya, Jingnan Bao, and Yilei Zhang* School of Mechanical and Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, N3.2-02-65, Singapore 639798, Singapore S Supporting Information *
ABSTRACT: Studies on the mammalian brain cerebral cortex have gained increasing importance due to the relevance of the region in controlling critical higher brain functions. Interactions between the cortical cells and surface extracellular matrix (ECM) proteins play a pivotal role in promoting stable cell adhesion, growth, and function. Poly(dimethylsiloxane) (PDMS) based platforms have been increasingly used for on-chip in vitro cellular system analysis. However, the inherent hydrophobicity of the PDMS surface has been unfavorable for any long-term cell system investigations due to transitory physical adsorption of ECM proteins on PDMS surfaces followed by eventual cell dislodgement due to poor anchorage and viability. To address this critical issue, we employed the (3-aminopropyl)triethoxysilane (APTES) based crosslinking strategy to stabilize ECM protein immobilization on PDMS. The efficiency of surface modification in supporting adhesion and long-term viability of neuronal and glial cells was analyzed. The chemically modified surfaces showed a relatively higher cell survival with an increased neurite length and neurite branching. These changes were understood in terms of an increase in surface hydrophilicity, protein stability, and cell−ECM protein interactions. The modification strategy could be successfully applied for stable cortical cell culture on the PDMS microchip for up to 3 weeks in vitro. KEYWORDS: cerebral cortex, neurons, glial cells, (3-aminopropyl)triethoxysilane (APTES), poly(dimethylsiloxane), protein immobilization
■
INTRODUCTION Miniaturized cell culture systems offer precise control of cell numbers in a given area or volume for minimalistic as well as single cell investigations, with minimal reagent consumption and cell loss.1−3 In recent years, poly dimethylsiloxane (PDMS) has been one of the predominant materials in microfabrications of miniaturized study platforms of several cellular processes due to its ease of fabrication, low cost, and optical transparency.4−6 Furthermore, the flexibility and versatility of tunable mechanical and surface topographical6 features up to submicron precision has increasingly facilitated rapid-prototyping of cell systems at the exploratory stages of fundamental research.4,6,7 These advantages have been tapped for in vitro study platforms of cell differentiation,8 tumor progression,9 cell−cell/environment interactions,10−12 physiology,13 functional studies,4,14 and disease models.15,16 In most of these studies, in vivo microarchitectures have been mimicked at cellular, tissue, and organ level architectures on chip. Of recent interest is to explore complex mechanisms involved in neural stem cell differentiation,17 disease progression,18 high−throughput drug screening,19 and cell−cell interactions of brain20 on PDMS microsystems. In this regard, there has been a growing interest © 2015 American Chemical Society
in the study of primary neural cell cultures from cerebral cortex digest for detailed studies on specific physiological behaviors such as axonal guidance,21 cell maturation,22 cell-surfacechemical/topographical interactions,23 and synaptogenesis24 under different physiologically relevant conditions and microenvironments. Moreover, the vital functional roles played by the cerebral cortex in memory, higher thinking, and sensory and motor perception have been studied.24 Neurodegenerative disorders such as Alzheimer’s disease, Parkinson’s disease, and amyotrophic lateral sclerosis have been linked to the changes in biochemical and cellular microenvironments during aging. In all these studies, the high inaccessibility to the brain makes it desirable to study neuronal function and degeneration on an appropriate in vitro model system, which requires one to closely mimic the in vivo state with high physiological relevance.25 Such in vitro models would enhance drug screening and biomarker discovery.26,27 The study of the cerebral cortex cell functions on PDMS microchips with suitable design and microenvironmenReceived: September 24, 2015 Accepted: October 27, 2015 Published: October 27, 2015 25529
DOI: 10.1021/acsami.5b09032 ACS Appl. Mater. Interfaces 2015, 7, 25529−25538
Research Article
ACS Applied Materials & Interfaces
Figure 1. Schematic representation of PDMS surface modification for the cerebral cortex cell culture.
influence on adhesion and behavior of the brain cerebral cortex cells in vitro are not known. Foreseeing a need for a stable and functional cortical culture system, in this work, we employed the aminosilane functionalization using (3-aminopropyl)triethoxysilane (APTES) for the immobilization of fibronectin, collagen type IV, poly-L-lysine, and laminin proteins (Figure 1). These proteins have been commonly studied as key neural cell extracellular matrix (ECM) proteins influencing the cortex cell−microenvironment interactions and as compatible substrate platforms for in vitro cell studies.48−52 The proteins interact with charged NH2 domains of APTES by ionic interactions.53 Subsequently, we perform primary culture of brain cells dissected and digested from the postnatal (P1) rat cerebral cortex on the ECM protein biofunctionalized substrates and assess the adhesion, survival, and morphology of neural and glial cell types until 21 DIV. We further test the implementation of this strategy to modify the microchannels of a widely used PDMS microchip with compartmentalized configuration which is a highly adaptable platform to study axonal growth, injury and regeneration, and cell−cell interactions, as well as to unravel the chemical and biomolecular signatures of neurons and glial cells.54,58,63 Prolonged analysis of cortical cells in such microsystems could eventually provide deeper insights into the interplay between extracellular signals and the consequent intracellular responses under various conditions concerning the cerebral cortex development and degeneration.
tal considerations could offer higher specificity and control with minimal cell and reagent consumption. However, one of the discouraging factors concerning the use of PDMS systems is its low biocompatibility due to inherent high surface hydrophobicity.28−30 Attempts have been made to reduce surface hydrophobicity of PDMS.31−33 Although this problem could be averted for a short time period by either protein absorption or plasma treatment, these effects are transient due to hydrophobic recovery32,33 leading to eventual detachment and clumping and death of cells within 1 week of cell seeding.28,34 Such problems caused by the short-term and weaker surface functionalization34−37 poses serious limitations on any long-term in vitro studies concerning synaptogenesis,38 neural aging and death,39 synaptic plasticity,40 and neural information processing41 on PDMS surfaces. Synaptic proteins such as synaptophysin and β-synuclein as well as the glutamate receptors in the cultured neurons reach their maximum expression level starting from 15 to 20 days after cell plating.39 Moreover, β-synucleins and β-amyloids which are some of the key factors in neurodegenerative disease (e.g., Parkinson’s and Alzheimer’s diseases) progression are known to localize within mature neurons upon growth for 20 days in vitro (DIV).39 Furthermore, the in vitro neural network matures with culture age and starts to exhibit full electrical activity only after 2 weeks of cell plating.42,43 This indicates that maintaining a long-term culture of primary cortical cells is important for a meaningful investigation of cell growth and maturation, neuronal network formation, signaling, etc. Interactions between the proteins and the substrate surfaces are often dominated by weak44 forces such as the electrostatic, hydrophobic, and van der Waals interactions which could cause gradual protein leaching into the culture medium. As a result, the matrix protein support on the PDMS surface eventually loses its biocompatibility to promote cell culture. Previous studies have attempted to reduce such effects by introducing stable and strong covalent binding between the proteins and material surfaces.45−47 Recently, we had shown that covalent cross-linking of extracellular matrix (ECM) proteins could promote enhanced mesenchymal stem cell adhesion, compared to traditional protein adsorption techniques. The resulting PDMS surfaces, with reduced hydrophobicity, favored stable and long-term anchorage and differentiation of cells.47,8 However, to our knowledge, the effect of such modification chemistry on the stability of ECM protein binding and eventual
■
METHODS
Modification of Planar PDMS Surface. Silicone elastomer base and the curing agent (SYLGARD, Dow Corning, USA, MW ∼15 000 Da at a kinematic viscosity of 3398.05 mm2/s) were mixed in the ratio of 10:1 and cast onto either a flat polystyrene well plate or dish. PDMS was then subjected to degassing for 30 min and cured at 70 °C for 90 min in a vacuum oven. The substrates were allowed to cool at room temperature and split into four groups based on the proteins: fibronectin (Life Technologies, Singapore), collagen IV (Life Technologies, Singapore), laminin (Life Technologies, Singapore), and poly-L-lysine (Life Technologies, Singapore). The cured PDMS substrates, in groups, were treated with oxygen plasma for 3 min (Harrick Plasma-PDC 32G) and subsequently immersed in 10% APTES (v/v in water) (Sigma-Aldrich, Singapore) at 54 °C for 120 min. Following APTES treatment, the substrates were thoroughly washed in nuclease-free water (Life Technologies, Singapore) and immersed in 20 μg/mL of one of the four proteins and stored along with unmodified substrates which were used as negative controls. 25530
DOI: 10.1021/acsami.5b09032 ACS Appl. Mater. Interfaces 2015, 7, 25529−25538
Research Article
ACS Applied Materials & Interfaces
(thickness ∼170 μm) at the bottom. Immediately following the plasma bonding, the surface modification of the microchannels was initiated by APTES and protein treatment in similar conditions as described in the modification of planar substrates. Chemical treatment and washing steps were performed by maintaining optimal hydrodynamic flows of the reagents/washing buffers at the inlet and exit reservoir ports of the microchip. Immunostaining with Neural and Glial Cell Markers. The substrates were rinsed with PBS (1×) and fixed in 3% glutaraldehyde solution for 20 min at room temperature. Cells were permeabilized for 15 min in 0.1% Triton X-100 (Sigma-Aldrich, Singapore) solution and treated for 30 min with 1% bovine serum albumin (BSA) (Life Technologies, Singapore) in PBS to block any nonspecific binding of the antibodies. Then, the cells were incubated with neuron-specific Microtubule Associated Protein-2 (MAP2) (Millipore, Singapore) and glial specific Glial Fibrillary Acidic protein (GFAP) (Millipore, Singapore) in 1:1000 dilution and stored overnight at 4 °C. On the following day, the cells were incubated with Alexafluor 488 and Alexafluor 555 secondary antibodies (Life Technologies, Singapore) against the neuronal and glial primary antibodies and stored overnight at 4 °C. Finally, the cells were also stained with nuclear DAPI (Life Technologies, Singapore) dye for 5 min. Neurite Analysis. Neurite analysis was performed using the ImagePro plus image analysis tool (Media Cybernetics Inc. Rockville, USA). Neurons were imaged in a bright-field microscope at 10× magnification on images collected on the 3rd, 7th, 13th, and 21st days in vitro. After image collection, image contrast was adjusted on the software to threshold coloration and contrast for accurate analysis of neurite details such as primary branching and neurite length. These data were exported to MS Excel for tabulation and analysis. Statistical Analysis. To determine the statistical significance in measurements between the compared data groups, the Student’s t test (unpaired) was adopted using Minitab 16 Statistical Software (Minitab Inc., USA) with 95% confidence intervals. All experiments were performed in quadruplicate or 4 well modified/unmodified substrates (for cortex cell culture). For each group, culture images for measurement of neurite length and branching were taken for at least 4 regions on each well consisting of at least 20 data points per image. A p-value of 10 MO) of a 5 μL volume was dispensed onto the PDMS substrate surfaces for drop shape analysis using OneAttension software (Attension, Finland). The angles formed at the interface of the droplet and substrate surface were measured by the static sessile drop (tangent) method for at least three distinct contact points per sample. Analysis of Elemental Composition. Elemental composition of the APTES modified and native PDMS substrates was determined to ascertain the presence of reactive functional groups introduced by chemical modification. Substrates were loaded onto the sample chamber maintained at 9 to 10 mbar pressure, and the spectra were collected under normal emission within 10 min. C 1s peaks, which correspond to hydrocarbons, were calibrated to a binding energy of 284.8 eV to account for the energy shift induced by charging. While maintaining the empirical Wagner sensitivity factors, 50 eV pass energy was used to obtain C 1s and N 1s specific region scans on the surface. Surface Protein Stability. The amount of protein retained on APTES+protein and unmodified/protein adsorbed PDMS substrates was determined by micro-BCA Protein Assay (Thermo Scientific, Singapore). Native PDMS was used as control substrate. Following the overnight incubation of each protein solution (20 μg/mL at 4 °C), the protein solutions were pipetted off. The substrates were stored under cell incubation conditions (5% CO2 at 37 °C in sterilized 1× PBS buffer, pH 7.2) for 21 days to mimic cell culture conditions on the similar protein immobilized substrates on which cells were seeded. Substrates were then treated with 0.05% Tween 20 (Sigma-Aldrich, Singapore) on a rotating shaker for 30 min and washed twice with nuclease free water to flush out any loosely immobilized and nonspecifically bound protein molecules. Protein adhesion on the substrate surfaces was determined as per the suggested Micro BCA standard protocol. Absorbance was measured at 562 nm on a Multiskan GO microplate spectrophotometer (Thermo Scientific, Singapore). Finally, the percentage ratio of the concentration of surface-bound protein to the initial protein seeding concentration was calculated from the absorbance correlation to reflect the percentage of protein retained on the PDMS surfaces. Cerebral Cortex Primary Cell Culture. Cortical tissues were isolated from the cerebral cortex region of the postnatal 1 day old (P1) Sprague−Dawley rat (Invivos, Singapore) brain. All animal works were approved by the Institutional Animal Care and Use Committee (IACUC) at Nanyang Technological University, Singapore, and duly abided by the guidelines for ethical treatment of animal specimen. The cortex region was carefully dissected in 1× PBS solution and dissociated by trituration after digestion with papain (20 μg/mL) (Life Technologies, Singapore) in a dissection buffer adjusted to neutral pH and allowed to digest until the formation of a smooth homogeneous precipitate. The cells were then suspended in a medium consisting of 50% minimal essential medium (MEM) supplemented with 10% fetal bovine serum and 50% Neurobasal supplemented with B27 and 0.5 mM glutamine, 25 μM glutamate (Life Technologies, Singapore), and 25 μM β-mercaptoethanol (Sigma-Aldrich, Singapore). Finally, the cells were then seeded onto the modified and unmodified PDMS substrates or the microchips at a concentration of 0.3 × 106 cells/mL. PDMS Microchip Preparation. PDMS microchips were fabricated on the basis of a sample compartmentalized design replicated on PDMS (10:1 base and curing agent ratio) from SU-8 masters with reference to previously studied methods.54,63 Initially, a thin layer of SU-8 (Microchem, Singapore) was spin coated to a depth of 30 μm. Patterning of the microfluidic design was performed by photolithography. A biopsy punch (Miltex, Singapore) of 4 mm inner diameter was used to produce the inlet and outlet reservoirs of the microdevice. Devices were then assembled by plasma bonding to a thin coated planar PDMS layer supported by a glass coverslip
■
RESULTS X-ray Photoelectron Spectroscopy (XPS). XPS characterization was performed to reveal the surface elemental composition (Figure 2). The relative loading of nitrogen (N 1s) (Figure 2A) and carbon (C 1s) (Figure 2B) was found to be higher in APTES modified PDMS surfaces in comparison with the native (unmodified) PDMS, which could be attributed to the presence of amino (−NH2) and triethyl (−C2H5) functional groups of APTES, respectively, on the modified substrates. Water Contact Angle Measurements. The surface wettability of the native PDMS and the chemically modified (with APTES+protein) and unmodified (without APTES modification and protein adsorption) PDMS was evaluated by measuring the mean water droplet contact angle of the static drop dispensed on the substrate surface (Figure 3). The lowest wettability was observed on the surfaces of the native PDMS which ranged within the hydrophobic region (110.09 ± 3.12°). A slight decrease in contact angles was measured on the protein adsorbed PDMS surfaces. However, the surface wettability could still be measured within a hydrophobic range of ∼90° to 100°. However, APTES modified PDMS surfaces with subsequent protein immobilization showed a significant reduction in contact angle at a relatively hydrophilic range for all the tested ECM proteins. 25531
DOI: 10.1021/acsami.5b09032 ACS Appl. Mater. Interfaces 2015, 7, 25529−25538
Research Article
ACS Applied Materials & Interfaces
Figure 4. Amount of proteins retained on the APTES modified and protein adsorbed PDMS surfaces, calculated from the micro BCA assay. The graph is represented as the mean ± SD. ∗ indicates p-values of
