- Email: [email protected]
Patterned substrates fabricated by a controlled freezing approach and biocompatibility evaluation by stem cells
Materials Science and Engineering C 49 (2015) 390–399
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Patterned substrates fabricated by a controlled freezing approach and biocompatibility evaluation by stem cells Lei Qian a,1, Adham Ahmed a, Laurence Glennon-Alty b, Yonghong Yang a,b, Patricia Murray b, Haifei Zhang a,⁎ a b
Department of Chemistry, University of Liverpool, Oxford Street, Liverpool L69 7ZD, UK Institute of Translational Medicine, University of Liverpool, Liverpool L69 3GE, UK
a r t i c l e
i n f o
Article history: Received 16 September 2014 Received in revised form 17 December 2014 Accepted 7 January 2015 Available online 9 January 2015 Keywords: Controlled freezing Surface patterns Mesenchymal stem cells Nanoparticles
a b s t r a c t Patterned substrates have been widely used in the studies investigating how to regulate cell growth and alignment. Such substrates may be fabricated by various techniques such as photolithography, soft lithography and microcontact printing. We report here a facile approach to fabricate aligned and grid surface patterns by a controlled freezing approach and further investigate their biocompatibility. The fabrication has been demonstrated with polymers (hydrophilic & hydrophobic), nanoparticles (organic & inorganic), or mixtures of these components. For the aligned surface patterns, the spacings between the patterned ridges can be tuned by varying the freezing rates. The biocompatibility of the substrates is evaluated by WST-8 viability tests with cell counting kit-8 (CCK-8) and by culturing with mouse mesenchymal stem cells (mMSCs). Three surface-patterned substrates (PLGA, PLGA nanospheres with chitosan, and silica colloids) are evaluated in more details to show that the mMSCs can grow alongside the aligned ridges while the cells grow randomly when plain glass slides are used as control. Further observations show that PLGA substrates undergo degradation, and are thus unsuitable for cell culture over the longer term. On the other hand, the PLGA-chitosan substrate and silica substrate were stable and could maintain mMSC alignment throughout the culture period. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Stem cells are able to self-renew and differentiate to produce specialized progeny, and thus have enormous potential for tissue engineering-based therapies [1]. For instance, mesenchymal stem cells (MSCs) have been used to generate cartilage for airway replacement [2], and bone tissue for the repair of long-bone defects [3]. However, in order to differentiate and function appropriately, it is very important for the cells to be correctly aligned under certain conditions [4,5]. Surface patterned substrates have been used to investigate the effect of surface topography on stem cell growth, alignment and differentiation. This is because the 2D culture systems can simplify the analysis when deconstructing the stem cell niche and assessing the effects of individual niche components on stem cell fate [6]. For example, a micropatterned bioresorbable poly(lactide-co-glycolide) (PLGA) substrate fabricated by soft lithography was found to control human MSC (hMSC) differentiation along different lineages [7]. It was also shown that scaffold alignment and optimized mechanical simulation were sufficient to drive MSC differentiation without the need for additional chemical stimuli [4]. A number of methods for directing cell alignment using patterned ⁎ Corresponding author. E-mail address: [email protected] (H. Zhang). 1 Current address: Key Laboratory for Liquid–Solid Structural Evolution and Processing of Materials, Shandong University, Jinan 250061, China.
http://dx.doi.org/10.1016/j.msec.2015.01.034 0928-4931/© 2015 Elsevier B.V. All rights reserved.
substrates have been described; for instance, a mould patterning method with solvent casting to form CeO2 nanoparticle lines within a PLGA film could induce the CeO2-dependent alignment of cardiac stem cells and MSCs [8]. Drug-laden PLGA microspheres patterned into grooves by Teflon chips could direct cellular alignment and osteogenic commitment of adipose-derived stem cells for bone regeneration [9]. There are different ways to fabricate surface patterns [10–12], including (i) self-assembly; (ii) breath figures as templates; (iii) laser or light irradiation; (iv) pre-formed patterns as templates; and (v) direct writing [13]. Self-assembly of block copolymers can produce patterns with lamellar, cylindrical or spherical structures based on microphase separation of dissimilar polymer chains into ordered domains [14,15]. Based on arrays of condensed water droplets, breath figure templating generates ordered porous films on substrates or 3D objects [16,17]. For these self-assembly approaches, the structures are formed as a result of self-assembly via selection of polymers & substrates and control of preparation conditions. It may be difficult to create desired complex patterns. Inkjet printing and dip-pen nanolithography are widely used to fabricate patterns with computer-controlled programs [13,18,19]. These direct writing techniques would require the use of dedicated facilities. The most commonly used approaches to fabricating surface patterned structures may be laser lithography [20] or photolithography [21–23]. The potential problems with photolithography are polymer degradation caused by photo-oxidation and lack of ability to form
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
large-area low-cost high-resolution patterns [24]. Soft lithography employs the pre-patterned structures and the fabrication process relies on replication and pattern transfer [25]. This technique generates well-defined patterns with controllable surface properties, providing a promising platform for patterning cells and bioactive molecules [26, 27]. Microcontact printing is also very useful for patterning chemicals and biological materials onto surfaces [28]. Both soft lithography and microcontact printing require pre-formed stamps or substrates which again are usually prepared by photolithography or microfabrication. We describe here an alternative approach to preparing surface patterns by controlled freezing and subsequent freeze-drying. Freezedrying is a versatile technique to fabricate 3D porous materials which are widely used as scaffolds for tissue engineering [29]. Particularly, directional freezing can be used to produce 3D aligned porous materials [30,31]. In this method, a temperature gradient is applied to a solution or suspension, which orientates the growth of ice crystals. Sublimation of ice by freeze-drying results in the formation of aligned porous structures. 3D porous scaffolds with anisotropic or aligned structures have been prepared by directional freezing to guide cell growth and alignment [32–34]. Although a previous report briefly mentioned that it was possible to produce 2D surface patterns with gold nanoparticles [35], there has been no study on fabrication of 2D surface patterns with various components and the use of such prepared patterns. Due to the wide use of patterned substrates in biological studies, in this study, we aim to present a versatile approach for preparation of surface patterned substrates (aligned and grid patterns) from different materials by directional freezing. The cell viability and the aligned growth of mMSCs with the substrates are evaluated. 2. Materials and methods 2.1. Solutions and nanoparticle suspensions All polymers were used as received and the relevant aqueous solutions with the desired concentrations were prepared in distilled water. In2O3, TiO2, and ZnO particles (all purchased from NanoTek®) were suspended in 0.1 wt.% aqueous poly(sodium acrylate) (PSA, Mw 2100, Sigma Aldrich) solution. Silica colloidal suspensions were diluted to the desired concentrations from the as-purchased Ludox silica colloidal suspensions (Sigma Aldrich). Polystyrene microspheres were prepared by surfactant-free suspension polymerization with average diameters around 450 nm and a solid content of ~10% [36]. They were then washed and dispersed in 5 wt.% poly(vinyl alcohol) (PVA, Mw 10K, Sigma Aldrich) solution at the concentration of 2.0 wt.%. Silica microspheres were prepared by a modified Stöber method [37] while PLGA (Resomer® RG 503H, Boehringer Ingelheim) nanospheres were prepared by a common oil-in-water emulsion evaporation method (see supporting information). 2.2. Fabrication of surface patterns A freezing stage controlled by a computer with software (FDCS freeze-drying system, Linkam Scientific Instrument Ltd.) was used to perform directional freezing on glass slides. 2 μl of the stock solutions or suspensions was deposited on glass slide with cover. The glass slide was placed on two temperature-controlled metal plates separated by a distance of 2 mm in the horizontal axis. Under the standard condition, one metal plate was set up at a temperature of −15 °C (− 6 °C for PLGAdioxane solutions) and the other plate at 20 °C. The glass slides were moved between the two plates at an accurate rate as controlled by the computer, in the range of 20–1000 μm/s. It should be noted that the slide moving rates were used to represent the freezing rate. The slide moving rate is proportional to the freezing rate and can be finely controlled to vary the freezing rate. After freezing, the frozen sample on glass slide was placed into liquid nitrogen and then transferred into a
391
Virtis AdVantage freeze dryer (Biopharma Process Systems) with shelf temperature at −10 °C for 48 h. To make the surface hydrophobic, the as-purchased glass slides were treated with chloro(dimethyl)octylsilane [38] and then used for PLGAdioxane solution processing. The surface patterns of silica particles were sintered in a furnace (Carbolite, CWF1500) at 400 °C for 6 h so that the patterns could be firmly stuck to the glass slide. The substrates with patterned silica colloids were thus treated before employed for mMSC growth. 2.3. mMSC cell culture The D1 clonal MSC line, derived from bone marrow of BALB/c mice, was purchased from ATCC (CRL-12424). For routine culture, the cells were maintained on tissue culture plastic in standard culture medium (high glucose Dulbecco's Modified Eagle Medium (DMEM) (Invitrogen) supplemented with 10% foetal calf serum (FCS) and 2 mM L-glutamine) at 37 °C and 5% CO2. The patterned substrates for cell culture were fabricated on circle glass slides with a diameter of 15 mm. The substrates were sterilized by UV light for 15 min on each side before culturing. For cell viability WST-8 test, the patterned substrates fabricated with silica colloids (Ludox HS-30), aqueous chitosan solution, and PLGAdioxane solution were placed in a Corning Costar 24-well plate. A tissue culture cover slip (Sarstedt) was used as a positive control while the background control was only the standard medium with no cells. All the samples were tested in triplet and the average values were used. 1 cm3 D1 mMSC cells suspension at the concentration of 10,000/cm3 was seeded in each well and incubated for 24 h. The medium was then aspirated and cell counting kit −8 (CCK-8, Sigma Aldrich) at a concentration of 20 μl in 200 μl medium was added into each well. After incubating for 3 h, 100 μl supernatant from each well was transferred into a 96-well plate and analysed by an Anthos Labtec LP 400 microplate reader at the wavelength of 450 nm. For the culture study, the mMSCs were seeded onto 15 mm diameter circle substrates in 24-well plates (Nunc) in 250 μl droplets. Medium was topped up to 0.5 cm3 after 1 day. The mMSCs were seeded at a density of 6000 cells per cm2 for 3 to 5 days in standard medium. At the end of the culture period, cells were fixed with 4% (w/v) paraformaldehyde and stained with phalloidin labelled with Alexa Fluor 488 (Invitrogen) to visualize the F-actin cytoskeleton. Images were captured under a Leica DM2500 fluorescence microscope (Leica, Heidelberg, Germany) using a DFC350FX camera. For confocal microscopy, a Zeiss LSM 510 confocal laser scanning system mounted on a Zeiss Axiovert 200M (Carl Zeiss, Germany) was used. 2.4. Characterization Surface pattern morphology was investigated using scanning electron microscopy (SEM, S-4800 Hitachi). The samples were coated with gold using a sputter-coater (EMITECH K550X) for 2 min at 25 mA. The spacing were measured based on the SEM images across a width ~0.5 mm (500 μm). The average values were used to plot the graph of spacings vs freezing rates. The height of the aligned ridges on the surface patterned substrates was measured from the SEM images. During the SEM imaging, the samples could be tilted to a certain angle, which helped to measure the ridge height. The SEM technique was also used to observe mMSC growth on the patterned substrates. After four days of culture, mMSCs on the PLGA-chitosan substrate were fixed, dried, and coated with Au before imaging. The in-situ directional freezing was investigated using Olympus CX41 microscope equipped with a digital camera and a computer-controlled freeze stage. The Fourier Transform Infra Red (FTIR) imaging was performed using a focal plane array (FPA) detector. Each spectrum corresponded to a specific position on the surface of the ATR crystal (diamond). By plotting the integrated absorbance of a specific spectral band in the mid-IR spectrum, it was possible to obtain the distribution and amount of a particular compound in
392
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
the different parts of the imaged area. A Varian imaging system was coupled with a Bio-Rad 60A FTIR spectrometer and an ATR-FTIR Golden Gate™ accessory (Specac Ltd., UK) positioned in the large sample compartment. The images were created using Varian Resolution Pro™ 4.0 software. 3. Results and discussion Directional freezing has been developed and used to produce 3D aligned porous materials [29–31]. Aqueous solutions or suspensions are normally used although organic solutions have also been employed to fabricate porous hydrophobic materials. However, the controlled freezing method has rarely been used to fabricate surface patterned structures, which are commonly produced by other techniques [11–28]. 3.1. Surface patterns containing hydrophilic polymers/silica colloids We firstly investigated a range of water-soluble polymers [29–31]. The chosen polymers were dissolved in water to make aqueous solutions at the concentration of 0.5 w/v%. PVA, poly(sodium 4styrenesulfonate) (PSS, Mw 70K, Sigma Aldrich), sodium alginate (Sigma Aldrich) and sodium carboxymethyl cellulose (NaCMC, Mw 250K, Sigma Aldrich) were first chosen because of their properties and potential applications. PVA is nonionic while the others are ionic polymers. Sodium alginate and NaCMC are natural polymers and can be easily crosslinked by metal ions such as Ca2+ and Fe3+. PSS is an anionic polymer which has biological applications and may also be carbonized to produce sulfur-doped porous carbon. Using the controlled freezing approach followed by freeze drying, dry aligned patterns on glass slides with spatial spacings around 30 μm were readily produced (Fig. S1). 3D aligned porous silica/PVA composites were prepared in a previous study [35]. The presence of PVA served two purposes: (1) to help stabilize the silica colloid suspension and (2) hold the silica colloids together after freeze drying [35]. Indeed, freeze-drying silica colloidal suspensions without added polymer produced either powders or very loose structures. However, 2D surface patterns directly from silica colloidal suspensions could be produced by freeze-drying on glass slides. Different types of Ludox silica colloids were processed to form aligned surface patterns (Fig. S2). The same colloidal suspensions were also stabilized with 0.5 wt.% NaCMC and processed (Fig. S3). With the addition of polymer, the pattern line spacings were generally smaller. Effect of the freezing rate on line spacings for the polymer/silica colloid system was further investigated with a model system of 5 wt.% HS-30 silica colloids in 5 wt.% PVA aqueous solution. Here, the slide moving rate is used as the freezing rate. Fig. 1A shows that the line spacings can be easily varied by changing the freezing rate in the range of 40–500 μm/s. In general, the smaller the freezing rate, the wider the line spacings; when the freezing rate increases, the line spacing decreases. In the range of freezing rates investigated here, aligned surface patterns were all produced. When the freezing rate was increased further, the particles started to be encapsulated, which agrees with the mechanism proposed by Deville et al. [39]. A similar study was carried out with the suspensions of 10 wt.% HS-30 silica colloids in 0.5 wt.% NaCMC. The same trend was observed, i.e., higher freezing rates led to smaller line spacings (Fig. S4). For a given colloidal suspension and/or polymer suspension, the freeze rate is decisive in the fate of the particles/molecules — whether excluded or encapsulated by the freezing front. As a general trend, the particles are encapsulated in the freezing ice front when the freezing rate is fast and they can be rejected from the freezing front at a slow freezing rate under which conditions the aligned porous structure may be formed [39]. There is a critical velocity present which can determine whether an ordered structure or randomly porous structure is produced. In the range of freezing rates investigated in Fig. 1A, these freeze rate should have been below the critical freezing velocity to generate the aligned patterns [30,39]. Larger ice crystals are formed when the freezing rate is low (allowing more time for ice crystals to growth
while limiting the nucleation process). This explains why the line spacings are bigger when the freezing rates are lower in Fig. 1A. The mechanism study about the controlled freezing has been based on hard spheres or ceramic particles. For the real colloidal suspensions in the presence of polymers, it is a very complex process. There are many factors that can influence the fate of the polymer/colloids during the freezing process, include particle shape & size, particle material, surface properties, type of polymer, concentrations, and solvent, etc. To show the versatility of our method, different types of colloids, polymers, nanospheres and microspheres, organic and aqueous solutions/suspensions have thus been investigated. The dry NaCMC/silica colloid pattern was further imaged by SEM (Fig. 1B). Individual colloids are not observed because the sizes of HS30 silica colloids are around 15 nm and are thus too small to be seen at this magnification (Fig. 1B inset). One concern is whether all the silica colloids were excluded from the freezing front, i.e., located in the ridge but not on the slide surface between the ridges. This is important when such aligned surface patterns are used to direct cell growth where the surface properties of the ridge and the substrate play important roles [7–9]. Because the silica colloids are very small and the glass slide is mainly made of silica as well, it is not trivial to distinguish the silica colloids from the glass substrate. Instead, we chose to use larger silica microspheres (average size ~ 1 μm) prepared by a modified Stöber method [37]. There are two assumptions with this choice: (1) the larger microspheres can be easily seen; and (2) it is more difficult to exclude larger spheres from the freezing front. In other words, if the larger microspheres can be excluded cleanly from the freezing front, it should be the case for smaller spheres or molecules. Fig. 1C shows the aligned surface pattern of silica microspheres/PVA. The assembled microspheres can be clearly seen on the ridge while only a few microspheres are found between the ridges. This demonstrates that under such freezing conditions, colloidal particles can be excluded from the freezing front and form aligned surface patterns. Between the parallel ridges, the substrate surface is exposed. This well-defined surface property/ chemical cue of the aligned pattern may be of great benefit in directing cell growth and alignment. Fig. 1B & C only show part of the aligned surface patterns. The long range order may be seen from the low magnification SEM images (Fig. S5) but they are still part of the pattern. The area of the surface patterns is at least 1 mm × 1 mm from SEM imaging at different parts. The study was further extended to organic and mixed colloidal suspensions. Polystyrene (PS) colloids (with average size around 450 nm) were prepared by surfactant-free emulsion polymerization [36]. Aqueous 2 wt.% PS colloidal suspension was directionally frozen on glass slides. As observed from optical microscope, an aligned surface pattern is formed (Fig. 2A). When the PS microspheres are stabilized in 5 wt.% PVA solution, narrower line spacings (or the width of orientated ice crystals) are formed (Fig. 2B). This is consistent with the observation for the silica/PVA system (Figs. S2 & S3). Aligned dry surface patterns were formed after freeze drying (Fig. 2C). The aggregated PS microspheres can be observed from the inset image. Mixed colloids, PS microspheres/silica colloids (Fig. 2D) and PLGA nanospheres/silica colloids (Fig. 2E) were suspended in 5 wt.% PVA solution. Aligned surface patterns were produced in both cases although the detailed morphology is slightly different (Fig. 2D versus E). The morphology of the ridges in Fig. 2D is different from the others. There are spikes from the rather flat ridges. The feature heights including the spikes are around 3 μm. The low magnification SEM images for Fig. 2D–E have shown the long range order of these surface patterns (Fig. S6). 3.2. Aligned & grid surface patterns for metal oxide nanoparticle system Commercially available In2O3 and TiO2 powders were suspended in 0.1 wt.% PSA aqueous solution by stirring and sonication, and were compared with a ZnO nanoparticle suspension re-dispersed in 0.1 wt.% PSA solution. The In2O3 and TiO2 suspensions were not highly stable.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
393
Fig. 1. (A) Graph showing line spacings of the freeze-dried patterns vary with the freezing rates. The suspension of 5 wt.% HS-30 silica colloids in 5 wt.% PVA aqueous solution was directly frozen. The scale bars are the same for the three optical images shown here. And the SEM images of two-dimensional aligned silica structures prepared from (B) 15 wt.% HS-30 silica colloids stabilized with 1.5 wt.% NaCMC (the inset showing a close-up view of the ridge and the ridge height is about 1 μm) and (C) 15 wt.% silica microspheres stabilized with 5 wt.% PVA (the ridge height around 10 μm and the diameters of the microspheres around 1 μm).
However, due to the rapid freezing process, it was possible to produce the aligned surface patterns after directional freezing and freeze drying on glass slides (Fig. 3A–C). By performing directional freezing twice along different directions using the respective colloidal suspensions, grid-structured patterns are fabricated (Fig. 3D–F). The grid structures could be made of different materials (Fig. 3D & E) or from the same material (Fig. 3F). 3.3. Aligned surface patterns containing PLGA PLGA solution in dioxane at 2.5 w/v% was directionally frozen on a modified hydrophobic glass slide. Dendritic ice crystal structure was observed by optical imaging (Fig. S7). After removing the frozen dioxane by freeze drying, the corresponding dry pattern (resembling an aligned ladder structure) was produced (Fig. 4A). Increasing the freezing rate resulted in denser pattern structures with narrower line spacings (Fig. 4B–D). For biological applications with PLGA, aqueous media are always employed. PLGA is a hydrophobic polymer. It can be difficult for the cell culture medium to wet a PLGA substrate. This can negatively influence the attachment of the cells to the PLGA substrate. Preparing PLGA nanospheres and then dispersing them in a water-insoluble hydrophilic matrix may be a solution to this problem. PLGA nanospheres were
prepared by an emulsion evaporation approach (Fig. S8). In the previous section, the surface pattern from the mixture of silica colloids and PLGA nanospheres in PVA was formed (Fig. 2E). However, PVA is soluble in water and thus is not suitable as a cell culture substrate. A biopolymer, chitosan, was selected for this purpose because it is insoluble in aqueous medium at neutral pH but is soluble in acidified water [40]. To produce the aligned surface patterns, PLGA nanospheres were suspended in 1 wt.% aqueous chitosan solution (acidified with acetic acid). The standard hydrophilic glass slide (i.e., not hydrophobilized) was used. After directional freezing and freeze drying, dry aligned surface patterned structures were formed. Fig. 5A–C show the surface patterns made from 5 wt.% PLGA nanospheres. The line spacing is in the order of 50 μm. Some PLGA nanospheres are observed between the aligned ridges (Fig. 5B) while most of the PLGA nanospheres are embedded in the chitosan matrix. Increasing the concentration of PLGA nanospheres resulted in a denser patterned structure with narrower line spacings (Fig. 5D). 3.4. FTIR imaging of the PLGA-chitosan pattern FTIR spectroscopic imaging is a chemical imaging technique that can map chemical compositions of a heterogeneous system and has been used broadly in material and biomedical fields [41]. Here this technique
394
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
Fig. 2. Optical images of directional freezing of 2 wt.% PS system (A) in water and (B) in 5 wt.% PVA aqueous solution. The dark area is for the unfrozen solution while the black lines are for aggregated PS microspheres and the white stripes are the orientated ice crystals. The SEM images of: (C) PVA-stabilized PS pattern prepared from 2.0 wt.% PS microspheres in 5 wt.% PVA, the inset image (scale bar 500 nm) showing the PS microspheres on the ridge, the ridge height around 2 μm; (D) PVA-stabilized PS and silica colloids pattern prepared from 2 wt.% PS microspheres in 5 wt.% HS-30 silica colloidal and 5 wt.% PVA solution, ridge height around 3 μm; and (E) PVA-stabilized PLGA and silica colloids pattern where the concentrations for each component are 5 wt.%, ridge height about 2 μm.
Fig. 3. Surface patterns fabricated from colloidal suspensions of In2O3 3 wt.% (A), ZnO 5 wt.%(B), TiO2 3 wt.% (C), ZnO/In2O3 grid (D), In2O3/TiO2 grid (E) and HS-30/NaCMC grid (F). All metal oxide particles are suspended in 0.1 wt.% PSA solution. HS-30 colloidal suspension is 10 wt.% in 0.5 wt.% NaCMC solution. Scale bars: 50 μm. Freezing condition: − 20 to 15 °C; 200 μm/s.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
395
Fig. 4. Fabrication of PLGA patterns. PLGA surface patterns prepared using 2.5 w/v% PLGA solutions and different freezing rates at (A) 40, (B) 200, (C) 300 and (D) 500 μm/s. Freezing condition: −6 to 15 °C. Scale bars: 50 μm.
is used to observe how the PLGA nanospheres are distributed in the aligned chitosan [42]. PLGA can be distinguished from chitosan molecules, which have a proxy structure of glycosaminoglycans, by the presence of C_O bonds in PLGA. FTIR imaging was thus performed based on the υ(C_O) absorption band at 1752 cm−1. Fig. 6A shows the optical image of the PLGA-chitosan pattern. Fig. 6B shows the chemical mapping based on absorption at 1752 cm−1. The red area indicates a high concentration of PLGA. It indicates that the majority of PLGA nanospheres are located within the aligned chitosan pattern. 3.5. Evaluation of biocompatibility by mMSC growth and alignment Surface morphology and chemical compositions can have a pronounced effect on stem cell morphology and differentiation [43,44]. For example, in a study on hydrogel with surface wrinkles, hMSCs were observed attached to lamellar wrinkles where they spread and conformed to the shape of the pattern, whereas when plated within hexagonal patterns, they did not spread and remained rounded [44]. Here, we wanted to investigate how the surface patterned substrates fabricated by the controlled freezing method could impact the growth of mMSCs. The cell viability of the substrates was firstly evaluated. Because all the substrates were fabricated from the components of silica colloids, chitosan, or PLGA, three substrates from each of them were evaluated by the WST-8 viability test (substrates 8 and 16 in Table S1 and the chitosan substrate fabricated by freezing rate 100 μm/s from
1 wt.% chitosan solution). Fig. 7 shows the viability of mMSCs for the substrates against a tissue culture cover slip for which the cell viability is set as 100%. The chitosan pattern exhibits a similar viability (104.61%) while the PLGA and silica substrates have lower viability (76.66% and 64.85%, respectively). A standard calibration curve was established by culturing mMSCs without substrates (Fig. S9). The cells grown on the substrates can then be calculated. When 10,000 cells were seeded in each well, the numbers of cells on the substrates after incubating 24 h were 11,300 for the chitosan substrate, 7600 for the PLGA substrate, and 6000 for the silica substrates, respectively. This demonstrates these substrates are biocompatible for mMSCs. It also shows the importance to prepare the patterned substrates from aqueous chitosan solution with suspended PLGA nanospheres. The substrates prepared under different conditions were screened for mMSCs culturing (Table S1). On all the substrates that were stable during cell culture, mMSCs survived and were able to align along distinct ridges in the substrate, while also demonstrating preference for areas of polymer over glass. It is noteworthy that the mMSCs aligned on the patterned surfaces containing silica. Although the area between the silica-based ridges is glass, which is also mainly composed of silica, it has been demonstrated previously that surface patterning can induce cell alignment, despite the substrate being chemically homogeneous [45]. As demonstrated by Fujita et al. [46], this is most likely due to the fact that the actin cytoskeleton forms more robust cell–substrate adhesions (i.e., focal adhesions) in cell protrusions that are parallel to the
396
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
Fig. 5. Aligned PLGA surface patterns fabricated from (A–C) 5% PLGA nanospheres in 1% aqueous chitosan solution. The ridge height around 8 μm; (D) 15% PLGA nanospheres in 1% aqueous chitosan solution. The ridge height around 5 μm. The area between the parallel chitosan ridges is the glass slide surface while the spheres are PLGA. Freezing condition: − 20 to 15 °C; 200 μm/s.
long axis of a topographical feature (such as a groove or ridge), than in cell protrusions that are perpendicular to the long axis; the result is that the actin cytoskeleton of the cell becomes aligned with the surface pattern. Several of the PLGA substrates degraded over the culture period, and thus had little effect on mMSC behaviour. Where polymer patterns remained intact, mMSCs commonly aligned parallel to the ridged structures within the pattern, though areas of non-aligned monolayers were also present in some areas on the majority of substrates, possibly due to over-growth of mMSCs. In contrast to PLGA, the PLGA-chitosan and silica patterned substrates created stable patterns and mMSCs were found to align with the ridges within the pattern. Following this initial observation (Table S1), the following substrates were selected from each set of substrates for observation
of mMSC growth: (i) substrate 8, a PLGA substrate (prepared from 10% PLGA-dioxane solution, temperature gradient − 6 °C to 15 °C, freezing rate 100 μm/s); (ii) substrate 11, a PLGA-chitosan substrate (prepared from 5 wt.% PLGA nanospheres in 1% chitosan solution, temperature gradient − 20 °C to 15 °C, freezing rate 100 μm/s); and (iii) substrate 16, a silica substrate (prepared from silica colloidal HS-30 suspension, temperature gradient − 20 °C to 15 °C, freezing rate 100 μm/s). mMSCs were seeded onto the PLGA, PLGA-chitosan, silica and glass (control) substrates. Following a 3 day culture period, the cells were fixed and stained with phalloidin to visualize the actin cytoskeleton. Because these surface patterns are formed on glass slides, the glass slide is used as a control rather than the commonly used polystyrene substrate. It is clear that all the surface patterns are able to align the growth of
Fig. 6. (A) Optical image of the PLGA nanospheres-chitosan pattern prepared from 15 wt.% PLGA nanospheres in 1% chitosan solution. Freezing condition: −20 to 15 °C; 100 μm/s. (B) FTIR image of the corresponding area at 1752 cm−1. The image area is 50 μm × 50 μm.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
104.61
100
Cell Viability (%)
100 76.66 64.84
50
0
silica
chitosan
PLGA
tissue culture slip
Fig. 7. The cell viability of the substrates with aligned patterned surface from silica colloids, chitosan, PLGA, The tests were performed with mMSCs and CKK-8 reagent, against a standard tissue culture cover slip.
mMSCs (Fig. 8). On the control glass substrate, the growth of mMSCs showed no evidence of alignment (Fig. 8A). On the PLGA patterned substrate, mMSCs formed a more complex arrangement, possibly due to the chemical composition of the substrate and/or the ladder-like aligned pattern of this substrate (Fig. 8B), whereas on the PLGA-chitosan and silica substrates, the cells displayed a spindle-shaped morphology and were aligned along the parallel ridge structures within the substrate (Fig. 8C and D). After culture for 5 days, confocal laser microscopy of phalloidin-stained cells confirmed the spindle shaped morphology and showed that the long axis of most cells was aligned parallel to the ridges within the PLGA-chitosan (Fig. S10) and silica substrates (Fig. S11). On the PLGA substrate, mMSCs appeared to reside in grooves between polymer ridges, and extended between comb-like gaps (Fig. S12).
397
Beyond their biocompatibility, the studies so far have demonstrated PLGA-chitosan and silica substrates are capable of directing the alignment of mMSC growth. The substrates can also be imaged by SEM at high magnification to visualize the location of mMSCs with the substrate. Due to the fact that the PLGA nanospheres can be easily detected under SEM, cells cultured for four days on PLGA-chitosan substrates were selected for SEM imaging. Figs. 9 and S13 show that the mMSCs are aligned parallel to ridged structures. The spindle-shaped cells are observed on the pattern ridges or in the region of the glass slide, but, in the latter case, the cells appear to extend filopodia to contact regions of polymer nanospheres. It has recently been shown that while lamellipodia mediate cell attachment and spreading on flat surfaces, filopodia play a key role in cell attachment to surfaces with nanotopographical features. Although the mechanisms have not been fully elucidated, it has been suggested that filopodia are unable to form stable adhesions on rigid, flat substrates, such as glass, and are rapidly replaced by lamellipodia, whereas on less rigid surfaces with nanotopographical features, the filopodia can form more stable adhesions, allowing the cell to apply the traction forces that are required for alignment [47,48]. Various scaffolds or substrates have been used to orientate the growth of stem cells [4–9,44–46]. With the newly prepared surface patterned substrates by the controlled freezing approach, this study with the mMSCs demonstrates the biocompatibility and the potential to direct cell growth. The limited issue for the current substrates is the stability of the patterned structures during cell culture. It had been observed that for some of the substrates the patterned structure could float off the glass cover slip in culture. This may be improved by varying the hydrophilicity and/or surface functionality of the glass surface or by use of other types of substrates (e.g., polystyrene, mica) to enhance the interaction of the patterned component and the substrate. Further study is required to understand the behaviour of mMSCs with the patterned
Fig. 8. Fluorescent microscopic images of the actin cytoskeleton of mMSCs on freeze-dried substrates following a 3 day culture period: plain glass (A), PLGA substrate (B), PLGA-chitosan substrate (C) and silica substrate (D). mMSCs were stained with phalloidin to visualize the actin cytoskeleton (green) and DAPI to visualize cell nuclei (blue). The actin cytoskeletons can be seen aligning with the patterned substrates. The same scale bars for all the images.
398
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
evidence for the aligned growth of mMSCs on the PLGA-chitosan substrate. This study demonstrates that a versatile controlled freezing approach can be used to fabricate various patterned substrates that have the potential to be used for guiding stem cell growth. Acknowledgements The authors are grateful for the financial support by the EPSRC (EP/ F016883/1) and access to the facility in the Centre for Materials Discovery at the University of Liverpool. The authors also gratefully acknowledge the financial support from a project grant awarded by the BBSRC (BB/D014638/1). We thank Prof. S. Kazarian group of Imperial College London for FTIR imaging. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2015.01.034. References
Fig. 9. SEM images of mMSCs attached and aligned to PLGA-chitosan substrate after culture for 4 days. The aligned mMSCs are parallel to ridges in a common direction, displaying a spindle-shaped morphology (A & B). In image B, the cells can be observed extending filopodia to contact regions of polymer (dashed line circles) whereas on glass, attachment is mediated by broader cell protrusions that resemble lamellipodia (solid line circle).
substrates and control of orientation angle of the aligned cells with the parameters including cell population & alignment angle, topology effect (e.g., groove width, ridge height, surface roughness), and substrate stiffness (e.g., by using of different nanoparticles and polymers). 4. Conclusions We have described a simple and versatile method to fabricate 2D surface patterned substrates. By spreading solutions or suspensions onto glass slides and directionally freezing using a computercontrolled freezing stage, followed by freeze drying, aligned surface patterns are readily produced. The preparation of aligned surface patterns for hydrophilic polymers, silica colloids, polymer/silica colloids, polymer colloids, and polymer/metal oxide nanoparticles, has been demonstrated. When the directional freezing procedures are applied twice in orthogonal directions, the substrates with grid surface patterns can be produced. The formation of such patterns is the result of orientated ice templating. The spacings between the aligned ridges may be tuned simply by varying freezing rates. In general, fast freezing rates lead to narrower spacings of the aligned patterns. The biocompatibility of the surface patterned substrates is demonstrated by WST-8 test and culturing with mMSCs. The chitosan pattern shows the best biocompatibility, very similar to the standard tissue culture slip. Compared to the plain glass slide where mMSC grow randomly, the aligned substrates (PLGA, PLGA nanospheres with chitosan, and silica colloids) can support and orientate the mMSC growth. While the stability and degradation of PLGA is an issue, the patterns fabricated with both PLGA-chitosan and silica colloid substrates are stable showing good performance in guiding mMSC alignment. The patterned substrate prepared from chitosan solution with suspended PLGA nanospheres gives better cell alignment and stability. SEM imaging provides further
[1] G.T. Daley, D.T. Scadden, Cell 132 (2008) 544–548. [2] P. Macchiarini, P. Jungebluth, T. Go, M.A. Asnaghi, L.E. Rees, T.A. Cogan, A. Dodson, J. Martorell, S. Bellini, P.P. Parnigotto, S.C. Dickinson, A.P. Hollander, S. Mantero, M.T. Conconi, M.A. Birchall, Lancet 372 (2008) 2023–2030. [3] M. Manassero, V. Viateau, M. Deschepper, K. Oudina, D. Logeart-Avramoglou, H. Petite, M. Bensidhoum, Tissue Eng. A 19 (2013) 1554–1563. [4] S.D. Subramony, B.R. Dargis, M. Castillo, E.U. Azeloglu, M.S. Tracey, A. Su, H.H. Lu, Biomaterials 34 (2013) 1942–1953. [5] M. Théry, J. Cell Sci. 123 (2010) 4201–4213. [6] M.P. Lutolf, P.M. Gilbert, H.M. Blau, Nature 462 (2009) 433–441. [7] C.Y. Tay, M. Pal, H. Yu, W.S. Leong, N.S. Tan, K.W. Ng, S. Venkatraman, F. Boey, D.T. Leong, L.P. Tan, Small 7 (2011) 1416–1421. [8] C. Mandoli, F. Pagliari, S. Pagliari, G. Forte, P.D. Nardo, S. Licoccia, E. Traversa, Adv. Funct. Mater. 20 (2010) 1617–1642. [9] X. Shi, S. Chen, J. Zhou, H. Yu, L. Li, H. Wu, Adv. Funct. Mater. 22 (2012) 3799–3807. [10] S.W. Hong, J. Huh, X. Gu, D.H. Lee, W.H. Jo, S. Park, T. Xu, T.P. Russell, PNAS 109 (2012) 1402–1406. [11] A.M. Hung, H. Noh, J.N. Cha, Nanoscale 2 (2010) 2530–2537. [12] C.M. Kolodziej, H.D. Maynard, Chem. Mater. 24 (2012) 774–780. [13] Y. Xu, F. Zhang, X. Feng, Small 7 (2011) 1338–1360. [14] H.C. Kim, S.M. Park, W.D. Hinsberg, Chem. Rev. 110 (2010) 146–177. [15] I.W. Hamley, Prog. Polym. Sci. 34 (2009) 1161–1210. [16] M.H. Stenzel, C. Barner-Kowollik, T.P. Davis, J. Polym. Sci. A Polym. Chem. 44 (2006) 2363–2375. [17] U.H.F. Bunz, Adv. Mater. 18 (2006) 973–989. [18] R.D. Piner, J. Zhu, F. Xu, S. Hong, C.A. Mirkin, Science 283 (1999) 661–663. [19] C.A. Mirkin, ACS Nano 1 (2007) 79–83. [20] J. Yamaguchi, K. Itaka, T. Hayakawa, K. Arai, M. Yamashiro, S. Yaginuma, H. Koinuma, Macromol. Rapid Commun. 25 (2004) 334–338. [21] W. He, C.R. Halberstadt, K.E. Gonsalves, Biomaterials 25 (2004) 2055–2063. [22] Z. Nie, E. Kumacheva, Nat. Mater. 7 (2008) 277–290. [23] P. Chandra, K. Lai, H.J. Sung, N. Murthy, J. Kohn, Biointerphases 5 (2010) 53–59. [24] Y. Xia, J.A. Rogers, K.E. Paul, G.M. Whitesides, Chem. Rev. 99 (1999) 1823–1848. [25] B.D. Gates, Q. Xu, J.C. Love, D.B. Wolfe, G.M. Whitesides, Annu. Rev. Mater. Res. 34 (2004) 339–372. [26] Y. Lu, S.C. Chen, Adv. Drug Deliv. Rev. 56 (2004) 1621–1633. [27] V.A. Schulte, Y. Hu, M. Diez, D. Bünger, M. Möller, M.C. Lensen, Biomaterials 31 (2010) 8583–8595. [28] S.A. Ruiz, C.S. Chen, Soft Matter 3 (2007) 168–177. [29] M. Darder, P. Aranda, M.L. Ferrer, M.C. Gutierrez, F. del Monte, E. Ruiz-Hitzky, Adv. Mater. 23 (2011) 5262–5267. [30] L. Qian, H. Zhang, J. Chem. Technol. Biotechnol. 86 (2011) 172–184. [31] S. Deville, Adv. Eng. Mater. 10 (2008) 155–169. [32] S.R. Caliari, B.A.C. Harley, Biomaterials 32 (2011) 5330–5340. [33] Q. Zhang, Y. Zhao, S. Yan, Y. Yang, H. Zhao, M. Li, S. Lu, D.L. Kaplan, Acta Biomater. 8 (2012) 2628–2638. [34] B.W. Riblett, N.L. Francis, M.A. Wheatley, U.G.K. Wegst, Adv. Funct. Mater. 22 (2012) 4920–4923. [35] H. Zhang, I. Hussain, M. Brust, M.F. Butler, S.P. Rannard, A.I. Cooper, Nat. Mater. 4 (2005) 787–793. [36] H. Zhang, J.Y. Lee, A. Ahmed, I. Hussain, A.I. Cooper, Angew. Chem. Int. Ed. 47 (2008) 4573–4576. [37] A. Ahmed, R. Clowes, E. Willneff, H. Ritchie, P. Myers, H. Zhang, Ind. Eng. Chem. Res. 49 (2010) 602–608. [38] A. Ahmed, W. Abdelmagid, H. Ritchie, P. Myers, H. Zhang, J. Chromatogr. A 1270 (2012) 194–203. [39] S. Deville, E. Maire, G. Bernard-Granger, A. Lasalle, A. Bogner, C. Gauthier, J. Leloup, C. Guizard, Nat. Mater. 8 (2009) 966–972. [40] L. Qian, H. Zhang, Green Chem. 12 (2010) 1207–1214.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399 [41] S.G. Kazarian, K.L.A. Chan, Analyst 138 (2013) 1940–1952. [42] S.G. Kazarian, K.L.A. Chan, V. Maquet, A.R. Boccaccini, Biomaterials 25 (2004) 3931–3938. [43] L. Glennon-Alty, R. Williams, S. Dixon, P. Murray, Acta Biomater. 9 (2013) 6041–6051. [44] M. Guvendiren, J.A. Burdick, Biomaterials 31 (2010) 6511–6518.
399
[45] S. Lenhert, M.B. Meier, U. Meyer, L. Chi, H.P. Wiesmann, Biomaterials 26 (2005) 563–570. [46] S. Fujita, M. Ohshima, H. Iwata, J. R. Soc. Interface 6 (2009) S269–S277. [47] J. Albushies, V. Vogel, Sci. Rep. 3 (2013) 1658. [48] B. Hoffmann, C. Schäfer, Cell Adhes. Migr. 4 (2010) 190–193.
Contents lists available at ScienceDirect
Materials Science and Engineering C journal homepage: www.elsevier.com/locate/msec
Patterned substrates fabricated by a controlled freezing approach and biocompatibility evaluation by stem cells Lei Qian a,1, Adham Ahmed a, Laurence Glennon-Alty b, Yonghong Yang a,b, Patricia Murray b, Haifei Zhang a,⁎ a b
Department of Chemistry, University of Liverpool, Oxford Street, Liverpool L69 7ZD, UK Institute of Translational Medicine, University of Liverpool, Liverpool L69 3GE, UK
a r t i c l e
i n f o
Article history: Received 16 September 2014 Received in revised form 17 December 2014 Accepted 7 January 2015 Available online 9 January 2015 Keywords: Controlled freezing Surface patterns Mesenchymal stem cells Nanoparticles
a b s t r a c t Patterned substrates have been widely used in the studies investigating how to regulate cell growth and alignment. Such substrates may be fabricated by various techniques such as photolithography, soft lithography and microcontact printing. We report here a facile approach to fabricate aligned and grid surface patterns by a controlled freezing approach and further investigate their biocompatibility. The fabrication has been demonstrated with polymers (hydrophilic & hydrophobic), nanoparticles (organic & inorganic), or mixtures of these components. For the aligned surface patterns, the spacings between the patterned ridges can be tuned by varying the freezing rates. The biocompatibility of the substrates is evaluated by WST-8 viability tests with cell counting kit-8 (CCK-8) and by culturing with mouse mesenchymal stem cells (mMSCs). Three surface-patterned substrates (PLGA, PLGA nanospheres with chitosan, and silica colloids) are evaluated in more details to show that the mMSCs can grow alongside the aligned ridges while the cells grow randomly when plain glass slides are used as control. Further observations show that PLGA substrates undergo degradation, and are thus unsuitable for cell culture over the longer term. On the other hand, the PLGA-chitosan substrate and silica substrate were stable and could maintain mMSC alignment throughout the culture period. © 2015 Elsevier B.V. All rights reserved.
1. Introduction Stem cells are able to self-renew and differentiate to produce specialized progeny, and thus have enormous potential for tissue engineering-based therapies [1]. For instance, mesenchymal stem cells (MSCs) have been used to generate cartilage for airway replacement [2], and bone tissue for the repair of long-bone defects [3]. However, in order to differentiate and function appropriately, it is very important for the cells to be correctly aligned under certain conditions [4,5]. Surface patterned substrates have been used to investigate the effect of surface topography on stem cell growth, alignment and differentiation. This is because the 2D culture systems can simplify the analysis when deconstructing the stem cell niche and assessing the effects of individual niche components on stem cell fate [6]. For example, a micropatterned bioresorbable poly(lactide-co-glycolide) (PLGA) substrate fabricated by soft lithography was found to control human MSC (hMSC) differentiation along different lineages [7]. It was also shown that scaffold alignment and optimized mechanical simulation were sufficient to drive MSC differentiation without the need for additional chemical stimuli [4]. A number of methods for directing cell alignment using patterned ⁎ Corresponding author. E-mail address: [email protected] (H. Zhang). 1 Current address: Key Laboratory for Liquid–Solid Structural Evolution and Processing of Materials, Shandong University, Jinan 250061, China.
http://dx.doi.org/10.1016/j.msec.2015.01.034 0928-4931/© 2015 Elsevier B.V. All rights reserved.
substrates have been described; for instance, a mould patterning method with solvent casting to form CeO2 nanoparticle lines within a PLGA film could induce the CeO2-dependent alignment of cardiac stem cells and MSCs [8]. Drug-laden PLGA microspheres patterned into grooves by Teflon chips could direct cellular alignment and osteogenic commitment of adipose-derived stem cells for bone regeneration [9]. There are different ways to fabricate surface patterns [10–12], including (i) self-assembly; (ii) breath figures as templates; (iii) laser or light irradiation; (iv) pre-formed patterns as templates; and (v) direct writing [13]. Self-assembly of block copolymers can produce patterns with lamellar, cylindrical or spherical structures based on microphase separation of dissimilar polymer chains into ordered domains [14,15]. Based on arrays of condensed water droplets, breath figure templating generates ordered porous films on substrates or 3D objects [16,17]. For these self-assembly approaches, the structures are formed as a result of self-assembly via selection of polymers & substrates and control of preparation conditions. It may be difficult to create desired complex patterns. Inkjet printing and dip-pen nanolithography are widely used to fabricate patterns with computer-controlled programs [13,18,19]. These direct writing techniques would require the use of dedicated facilities. The most commonly used approaches to fabricating surface patterned structures may be laser lithography [20] or photolithography [21–23]. The potential problems with photolithography are polymer degradation caused by photo-oxidation and lack of ability to form
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
large-area low-cost high-resolution patterns [24]. Soft lithography employs the pre-patterned structures and the fabrication process relies on replication and pattern transfer [25]. This technique generates well-defined patterns with controllable surface properties, providing a promising platform for patterning cells and bioactive molecules [26, 27]. Microcontact printing is also very useful for patterning chemicals and biological materials onto surfaces [28]. Both soft lithography and microcontact printing require pre-formed stamps or substrates which again are usually prepared by photolithography or microfabrication. We describe here an alternative approach to preparing surface patterns by controlled freezing and subsequent freeze-drying. Freezedrying is a versatile technique to fabricate 3D porous materials which are widely used as scaffolds for tissue engineering [29]. Particularly, directional freezing can be used to produce 3D aligned porous materials [30,31]. In this method, a temperature gradient is applied to a solution or suspension, which orientates the growth of ice crystals. Sublimation of ice by freeze-drying results in the formation of aligned porous structures. 3D porous scaffolds with anisotropic or aligned structures have been prepared by directional freezing to guide cell growth and alignment [32–34]. Although a previous report briefly mentioned that it was possible to produce 2D surface patterns with gold nanoparticles [35], there has been no study on fabrication of 2D surface patterns with various components and the use of such prepared patterns. Due to the wide use of patterned substrates in biological studies, in this study, we aim to present a versatile approach for preparation of surface patterned substrates (aligned and grid patterns) from different materials by directional freezing. The cell viability and the aligned growth of mMSCs with the substrates are evaluated. 2. Materials and methods 2.1. Solutions and nanoparticle suspensions All polymers were used as received and the relevant aqueous solutions with the desired concentrations were prepared in distilled water. In2O3, TiO2, and ZnO particles (all purchased from NanoTek®) were suspended in 0.1 wt.% aqueous poly(sodium acrylate) (PSA, Mw 2100, Sigma Aldrich) solution. Silica colloidal suspensions were diluted to the desired concentrations from the as-purchased Ludox silica colloidal suspensions (Sigma Aldrich). Polystyrene microspheres were prepared by surfactant-free suspension polymerization with average diameters around 450 nm and a solid content of ~10% [36]. They were then washed and dispersed in 5 wt.% poly(vinyl alcohol) (PVA, Mw 10K, Sigma Aldrich) solution at the concentration of 2.0 wt.%. Silica microspheres were prepared by a modified Stöber method [37] while PLGA (Resomer® RG 503H, Boehringer Ingelheim) nanospheres were prepared by a common oil-in-water emulsion evaporation method (see supporting information). 2.2. Fabrication of surface patterns A freezing stage controlled by a computer with software (FDCS freeze-drying system, Linkam Scientific Instrument Ltd.) was used to perform directional freezing on glass slides. 2 μl of the stock solutions or suspensions was deposited on glass slide with cover. The glass slide was placed on two temperature-controlled metal plates separated by a distance of 2 mm in the horizontal axis. Under the standard condition, one metal plate was set up at a temperature of −15 °C (− 6 °C for PLGAdioxane solutions) and the other plate at 20 °C. The glass slides were moved between the two plates at an accurate rate as controlled by the computer, in the range of 20–1000 μm/s. It should be noted that the slide moving rates were used to represent the freezing rate. The slide moving rate is proportional to the freezing rate and can be finely controlled to vary the freezing rate. After freezing, the frozen sample on glass slide was placed into liquid nitrogen and then transferred into a
391
Virtis AdVantage freeze dryer (Biopharma Process Systems) with shelf temperature at −10 °C for 48 h. To make the surface hydrophobic, the as-purchased glass slides were treated with chloro(dimethyl)octylsilane [38] and then used for PLGAdioxane solution processing. The surface patterns of silica particles were sintered in a furnace (Carbolite, CWF1500) at 400 °C for 6 h so that the patterns could be firmly stuck to the glass slide. The substrates with patterned silica colloids were thus treated before employed for mMSC growth. 2.3. mMSC cell culture The D1 clonal MSC line, derived from bone marrow of BALB/c mice, was purchased from ATCC (CRL-12424). For routine culture, the cells were maintained on tissue culture plastic in standard culture medium (high glucose Dulbecco's Modified Eagle Medium (DMEM) (Invitrogen) supplemented with 10% foetal calf serum (FCS) and 2 mM L-glutamine) at 37 °C and 5% CO2. The patterned substrates for cell culture were fabricated on circle glass slides with a diameter of 15 mm. The substrates were sterilized by UV light for 15 min on each side before culturing. For cell viability WST-8 test, the patterned substrates fabricated with silica colloids (Ludox HS-30), aqueous chitosan solution, and PLGAdioxane solution were placed in a Corning Costar 24-well plate. A tissue culture cover slip (Sarstedt) was used as a positive control while the background control was only the standard medium with no cells. All the samples were tested in triplet and the average values were used. 1 cm3 D1 mMSC cells suspension at the concentration of 10,000/cm3 was seeded in each well and incubated for 24 h. The medium was then aspirated and cell counting kit −8 (CCK-8, Sigma Aldrich) at a concentration of 20 μl in 200 μl medium was added into each well. After incubating for 3 h, 100 μl supernatant from each well was transferred into a 96-well plate and analysed by an Anthos Labtec LP 400 microplate reader at the wavelength of 450 nm. For the culture study, the mMSCs were seeded onto 15 mm diameter circle substrates in 24-well plates (Nunc) in 250 μl droplets. Medium was topped up to 0.5 cm3 after 1 day. The mMSCs were seeded at a density of 6000 cells per cm2 for 3 to 5 days in standard medium. At the end of the culture period, cells were fixed with 4% (w/v) paraformaldehyde and stained with phalloidin labelled with Alexa Fluor 488 (Invitrogen) to visualize the F-actin cytoskeleton. Images were captured under a Leica DM2500 fluorescence microscope (Leica, Heidelberg, Germany) using a DFC350FX camera. For confocal microscopy, a Zeiss LSM 510 confocal laser scanning system mounted on a Zeiss Axiovert 200M (Carl Zeiss, Germany) was used. 2.4. Characterization Surface pattern morphology was investigated using scanning electron microscopy (SEM, S-4800 Hitachi). The samples were coated with gold using a sputter-coater (EMITECH K550X) for 2 min at 25 mA. The spacing were measured based on the SEM images across a width ~0.5 mm (500 μm). The average values were used to plot the graph of spacings vs freezing rates. The height of the aligned ridges on the surface patterned substrates was measured from the SEM images. During the SEM imaging, the samples could be tilted to a certain angle, which helped to measure the ridge height. The SEM technique was also used to observe mMSC growth on the patterned substrates. After four days of culture, mMSCs on the PLGA-chitosan substrate were fixed, dried, and coated with Au before imaging. The in-situ directional freezing was investigated using Olympus CX41 microscope equipped with a digital camera and a computer-controlled freeze stage. The Fourier Transform Infra Red (FTIR) imaging was performed using a focal plane array (FPA) detector. Each spectrum corresponded to a specific position on the surface of the ATR crystal (diamond). By plotting the integrated absorbance of a specific spectral band in the mid-IR spectrum, it was possible to obtain the distribution and amount of a particular compound in
392
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
the different parts of the imaged area. A Varian imaging system was coupled with a Bio-Rad 60A FTIR spectrometer and an ATR-FTIR Golden Gate™ accessory (Specac Ltd., UK) positioned in the large sample compartment. The images were created using Varian Resolution Pro™ 4.0 software. 3. Results and discussion Directional freezing has been developed and used to produce 3D aligned porous materials [29–31]. Aqueous solutions or suspensions are normally used although organic solutions have also been employed to fabricate porous hydrophobic materials. However, the controlled freezing method has rarely been used to fabricate surface patterned structures, which are commonly produced by other techniques [11–28]. 3.1. Surface patterns containing hydrophilic polymers/silica colloids We firstly investigated a range of water-soluble polymers [29–31]. The chosen polymers were dissolved in water to make aqueous solutions at the concentration of 0.5 w/v%. PVA, poly(sodium 4styrenesulfonate) (PSS, Mw 70K, Sigma Aldrich), sodium alginate (Sigma Aldrich) and sodium carboxymethyl cellulose (NaCMC, Mw 250K, Sigma Aldrich) were first chosen because of their properties and potential applications. PVA is nonionic while the others are ionic polymers. Sodium alginate and NaCMC are natural polymers and can be easily crosslinked by metal ions such as Ca2+ and Fe3+. PSS is an anionic polymer which has biological applications and may also be carbonized to produce sulfur-doped porous carbon. Using the controlled freezing approach followed by freeze drying, dry aligned patterns on glass slides with spatial spacings around 30 μm were readily produced (Fig. S1). 3D aligned porous silica/PVA composites were prepared in a previous study [35]. The presence of PVA served two purposes: (1) to help stabilize the silica colloid suspension and (2) hold the silica colloids together after freeze drying [35]. Indeed, freeze-drying silica colloidal suspensions without added polymer produced either powders or very loose structures. However, 2D surface patterns directly from silica colloidal suspensions could be produced by freeze-drying on glass slides. Different types of Ludox silica colloids were processed to form aligned surface patterns (Fig. S2). The same colloidal suspensions were also stabilized with 0.5 wt.% NaCMC and processed (Fig. S3). With the addition of polymer, the pattern line spacings were generally smaller. Effect of the freezing rate on line spacings for the polymer/silica colloid system was further investigated with a model system of 5 wt.% HS-30 silica colloids in 5 wt.% PVA aqueous solution. Here, the slide moving rate is used as the freezing rate. Fig. 1A shows that the line spacings can be easily varied by changing the freezing rate in the range of 40–500 μm/s. In general, the smaller the freezing rate, the wider the line spacings; when the freezing rate increases, the line spacing decreases. In the range of freezing rates investigated here, aligned surface patterns were all produced. When the freezing rate was increased further, the particles started to be encapsulated, which agrees with the mechanism proposed by Deville et al. [39]. A similar study was carried out with the suspensions of 10 wt.% HS-30 silica colloids in 0.5 wt.% NaCMC. The same trend was observed, i.e., higher freezing rates led to smaller line spacings (Fig. S4). For a given colloidal suspension and/or polymer suspension, the freeze rate is decisive in the fate of the particles/molecules — whether excluded or encapsulated by the freezing front. As a general trend, the particles are encapsulated in the freezing ice front when the freezing rate is fast and they can be rejected from the freezing front at a slow freezing rate under which conditions the aligned porous structure may be formed [39]. There is a critical velocity present which can determine whether an ordered structure or randomly porous structure is produced. In the range of freezing rates investigated in Fig. 1A, these freeze rate should have been below the critical freezing velocity to generate the aligned patterns [30,39]. Larger ice crystals are formed when the freezing rate is low (allowing more time for ice crystals to growth
while limiting the nucleation process). This explains why the line spacings are bigger when the freezing rates are lower in Fig. 1A. The mechanism study about the controlled freezing has been based on hard spheres or ceramic particles. For the real colloidal suspensions in the presence of polymers, it is a very complex process. There are many factors that can influence the fate of the polymer/colloids during the freezing process, include particle shape & size, particle material, surface properties, type of polymer, concentrations, and solvent, etc. To show the versatility of our method, different types of colloids, polymers, nanospheres and microspheres, organic and aqueous solutions/suspensions have thus been investigated. The dry NaCMC/silica colloid pattern was further imaged by SEM (Fig. 1B). Individual colloids are not observed because the sizes of HS30 silica colloids are around 15 nm and are thus too small to be seen at this magnification (Fig. 1B inset). One concern is whether all the silica colloids were excluded from the freezing front, i.e., located in the ridge but not on the slide surface between the ridges. This is important when such aligned surface patterns are used to direct cell growth where the surface properties of the ridge and the substrate play important roles [7–9]. Because the silica colloids are very small and the glass slide is mainly made of silica as well, it is not trivial to distinguish the silica colloids from the glass substrate. Instead, we chose to use larger silica microspheres (average size ~ 1 μm) prepared by a modified Stöber method [37]. There are two assumptions with this choice: (1) the larger microspheres can be easily seen; and (2) it is more difficult to exclude larger spheres from the freezing front. In other words, if the larger microspheres can be excluded cleanly from the freezing front, it should be the case for smaller spheres or molecules. Fig. 1C shows the aligned surface pattern of silica microspheres/PVA. The assembled microspheres can be clearly seen on the ridge while only a few microspheres are found between the ridges. This demonstrates that under such freezing conditions, colloidal particles can be excluded from the freezing front and form aligned surface patterns. Between the parallel ridges, the substrate surface is exposed. This well-defined surface property/ chemical cue of the aligned pattern may be of great benefit in directing cell growth and alignment. Fig. 1B & C only show part of the aligned surface patterns. The long range order may be seen from the low magnification SEM images (Fig. S5) but they are still part of the pattern. The area of the surface patterns is at least 1 mm × 1 mm from SEM imaging at different parts. The study was further extended to organic and mixed colloidal suspensions. Polystyrene (PS) colloids (with average size around 450 nm) were prepared by surfactant-free emulsion polymerization [36]. Aqueous 2 wt.% PS colloidal suspension was directionally frozen on glass slides. As observed from optical microscope, an aligned surface pattern is formed (Fig. 2A). When the PS microspheres are stabilized in 5 wt.% PVA solution, narrower line spacings (or the width of orientated ice crystals) are formed (Fig. 2B). This is consistent with the observation for the silica/PVA system (Figs. S2 & S3). Aligned dry surface patterns were formed after freeze drying (Fig. 2C). The aggregated PS microspheres can be observed from the inset image. Mixed colloids, PS microspheres/silica colloids (Fig. 2D) and PLGA nanospheres/silica colloids (Fig. 2E) were suspended in 5 wt.% PVA solution. Aligned surface patterns were produced in both cases although the detailed morphology is slightly different (Fig. 2D versus E). The morphology of the ridges in Fig. 2D is different from the others. There are spikes from the rather flat ridges. The feature heights including the spikes are around 3 μm. The low magnification SEM images for Fig. 2D–E have shown the long range order of these surface patterns (Fig. S6). 3.2. Aligned & grid surface patterns for metal oxide nanoparticle system Commercially available In2O3 and TiO2 powders were suspended in 0.1 wt.% PSA aqueous solution by stirring and sonication, and were compared with a ZnO nanoparticle suspension re-dispersed in 0.1 wt.% PSA solution. The In2O3 and TiO2 suspensions were not highly stable.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
393
Fig. 1. (A) Graph showing line spacings of the freeze-dried patterns vary with the freezing rates. The suspension of 5 wt.% HS-30 silica colloids in 5 wt.% PVA aqueous solution was directly frozen. The scale bars are the same for the three optical images shown here. And the SEM images of two-dimensional aligned silica structures prepared from (B) 15 wt.% HS-30 silica colloids stabilized with 1.5 wt.% NaCMC (the inset showing a close-up view of the ridge and the ridge height is about 1 μm) and (C) 15 wt.% silica microspheres stabilized with 5 wt.% PVA (the ridge height around 10 μm and the diameters of the microspheres around 1 μm).
However, due to the rapid freezing process, it was possible to produce the aligned surface patterns after directional freezing and freeze drying on glass slides (Fig. 3A–C). By performing directional freezing twice along different directions using the respective colloidal suspensions, grid-structured patterns are fabricated (Fig. 3D–F). The grid structures could be made of different materials (Fig. 3D & E) or from the same material (Fig. 3F). 3.3. Aligned surface patterns containing PLGA PLGA solution in dioxane at 2.5 w/v% was directionally frozen on a modified hydrophobic glass slide. Dendritic ice crystal structure was observed by optical imaging (Fig. S7). After removing the frozen dioxane by freeze drying, the corresponding dry pattern (resembling an aligned ladder structure) was produced (Fig. 4A). Increasing the freezing rate resulted in denser pattern structures with narrower line spacings (Fig. 4B–D). For biological applications with PLGA, aqueous media are always employed. PLGA is a hydrophobic polymer. It can be difficult for the cell culture medium to wet a PLGA substrate. This can negatively influence the attachment of the cells to the PLGA substrate. Preparing PLGA nanospheres and then dispersing them in a water-insoluble hydrophilic matrix may be a solution to this problem. PLGA nanospheres were
prepared by an emulsion evaporation approach (Fig. S8). In the previous section, the surface pattern from the mixture of silica colloids and PLGA nanospheres in PVA was formed (Fig. 2E). However, PVA is soluble in water and thus is not suitable as a cell culture substrate. A biopolymer, chitosan, was selected for this purpose because it is insoluble in aqueous medium at neutral pH but is soluble in acidified water [40]. To produce the aligned surface patterns, PLGA nanospheres were suspended in 1 wt.% aqueous chitosan solution (acidified with acetic acid). The standard hydrophilic glass slide (i.e., not hydrophobilized) was used. After directional freezing and freeze drying, dry aligned surface patterned structures were formed. Fig. 5A–C show the surface patterns made from 5 wt.% PLGA nanospheres. The line spacing is in the order of 50 μm. Some PLGA nanospheres are observed between the aligned ridges (Fig. 5B) while most of the PLGA nanospheres are embedded in the chitosan matrix. Increasing the concentration of PLGA nanospheres resulted in a denser patterned structure with narrower line spacings (Fig. 5D). 3.4. FTIR imaging of the PLGA-chitosan pattern FTIR spectroscopic imaging is a chemical imaging technique that can map chemical compositions of a heterogeneous system and has been used broadly in material and biomedical fields [41]. Here this technique
394
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
Fig. 2. Optical images of directional freezing of 2 wt.% PS system (A) in water and (B) in 5 wt.% PVA aqueous solution. The dark area is for the unfrozen solution while the black lines are for aggregated PS microspheres and the white stripes are the orientated ice crystals. The SEM images of: (C) PVA-stabilized PS pattern prepared from 2.0 wt.% PS microspheres in 5 wt.% PVA, the inset image (scale bar 500 nm) showing the PS microspheres on the ridge, the ridge height around 2 μm; (D) PVA-stabilized PS and silica colloids pattern prepared from 2 wt.% PS microspheres in 5 wt.% HS-30 silica colloidal and 5 wt.% PVA solution, ridge height around 3 μm; and (E) PVA-stabilized PLGA and silica colloids pattern where the concentrations for each component are 5 wt.%, ridge height about 2 μm.
Fig. 3. Surface patterns fabricated from colloidal suspensions of In2O3 3 wt.% (A), ZnO 5 wt.%(B), TiO2 3 wt.% (C), ZnO/In2O3 grid (D), In2O3/TiO2 grid (E) and HS-30/NaCMC grid (F). All metal oxide particles are suspended in 0.1 wt.% PSA solution. HS-30 colloidal suspension is 10 wt.% in 0.5 wt.% NaCMC solution. Scale bars: 50 μm. Freezing condition: − 20 to 15 °C; 200 μm/s.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
395
Fig. 4. Fabrication of PLGA patterns. PLGA surface patterns prepared using 2.5 w/v% PLGA solutions and different freezing rates at (A) 40, (B) 200, (C) 300 and (D) 500 μm/s. Freezing condition: −6 to 15 °C. Scale bars: 50 μm.
is used to observe how the PLGA nanospheres are distributed in the aligned chitosan [42]. PLGA can be distinguished from chitosan molecules, which have a proxy structure of glycosaminoglycans, by the presence of C_O bonds in PLGA. FTIR imaging was thus performed based on the υ(C_O) absorption band at 1752 cm−1. Fig. 6A shows the optical image of the PLGA-chitosan pattern. Fig. 6B shows the chemical mapping based on absorption at 1752 cm−1. The red area indicates a high concentration of PLGA. It indicates that the majority of PLGA nanospheres are located within the aligned chitosan pattern. 3.5. Evaluation of biocompatibility by mMSC growth and alignment Surface morphology and chemical compositions can have a pronounced effect on stem cell morphology and differentiation [43,44]. For example, in a study on hydrogel with surface wrinkles, hMSCs were observed attached to lamellar wrinkles where they spread and conformed to the shape of the pattern, whereas when plated within hexagonal patterns, they did not spread and remained rounded [44]. Here, we wanted to investigate how the surface patterned substrates fabricated by the controlled freezing method could impact the growth of mMSCs. The cell viability of the substrates was firstly evaluated. Because all the substrates were fabricated from the components of silica colloids, chitosan, or PLGA, three substrates from each of them were evaluated by the WST-8 viability test (substrates 8 and 16 in Table S1 and the chitosan substrate fabricated by freezing rate 100 μm/s from
1 wt.% chitosan solution). Fig. 7 shows the viability of mMSCs for the substrates against a tissue culture cover slip for which the cell viability is set as 100%. The chitosan pattern exhibits a similar viability (104.61%) while the PLGA and silica substrates have lower viability (76.66% and 64.85%, respectively). A standard calibration curve was established by culturing mMSCs without substrates (Fig. S9). The cells grown on the substrates can then be calculated. When 10,000 cells were seeded in each well, the numbers of cells on the substrates after incubating 24 h were 11,300 for the chitosan substrate, 7600 for the PLGA substrate, and 6000 for the silica substrates, respectively. This demonstrates these substrates are biocompatible for mMSCs. It also shows the importance to prepare the patterned substrates from aqueous chitosan solution with suspended PLGA nanospheres. The substrates prepared under different conditions were screened for mMSCs culturing (Table S1). On all the substrates that were stable during cell culture, mMSCs survived and were able to align along distinct ridges in the substrate, while also demonstrating preference for areas of polymer over glass. It is noteworthy that the mMSCs aligned on the patterned surfaces containing silica. Although the area between the silica-based ridges is glass, which is also mainly composed of silica, it has been demonstrated previously that surface patterning can induce cell alignment, despite the substrate being chemically homogeneous [45]. As demonstrated by Fujita et al. [46], this is most likely due to the fact that the actin cytoskeleton forms more robust cell–substrate adhesions (i.e., focal adhesions) in cell protrusions that are parallel to the
396
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
Fig. 5. Aligned PLGA surface patterns fabricated from (A–C) 5% PLGA nanospheres in 1% aqueous chitosan solution. The ridge height around 8 μm; (D) 15% PLGA nanospheres in 1% aqueous chitosan solution. The ridge height around 5 μm. The area between the parallel chitosan ridges is the glass slide surface while the spheres are PLGA. Freezing condition: − 20 to 15 °C; 200 μm/s.
long axis of a topographical feature (such as a groove or ridge), than in cell protrusions that are perpendicular to the long axis; the result is that the actin cytoskeleton of the cell becomes aligned with the surface pattern. Several of the PLGA substrates degraded over the culture period, and thus had little effect on mMSC behaviour. Where polymer patterns remained intact, mMSCs commonly aligned parallel to the ridged structures within the pattern, though areas of non-aligned monolayers were also present in some areas on the majority of substrates, possibly due to over-growth of mMSCs. In contrast to PLGA, the PLGA-chitosan and silica patterned substrates created stable patterns and mMSCs were found to align with the ridges within the pattern. Following this initial observation (Table S1), the following substrates were selected from each set of substrates for observation
of mMSC growth: (i) substrate 8, a PLGA substrate (prepared from 10% PLGA-dioxane solution, temperature gradient − 6 °C to 15 °C, freezing rate 100 μm/s); (ii) substrate 11, a PLGA-chitosan substrate (prepared from 5 wt.% PLGA nanospheres in 1% chitosan solution, temperature gradient − 20 °C to 15 °C, freezing rate 100 μm/s); and (iii) substrate 16, a silica substrate (prepared from silica colloidal HS-30 suspension, temperature gradient − 20 °C to 15 °C, freezing rate 100 μm/s). mMSCs were seeded onto the PLGA, PLGA-chitosan, silica and glass (control) substrates. Following a 3 day culture period, the cells were fixed and stained with phalloidin to visualize the actin cytoskeleton. Because these surface patterns are formed on glass slides, the glass slide is used as a control rather than the commonly used polystyrene substrate. It is clear that all the surface patterns are able to align the growth of
Fig. 6. (A) Optical image of the PLGA nanospheres-chitosan pattern prepared from 15 wt.% PLGA nanospheres in 1% chitosan solution. Freezing condition: −20 to 15 °C; 100 μm/s. (B) FTIR image of the corresponding area at 1752 cm−1. The image area is 50 μm × 50 μm.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
104.61
100
Cell Viability (%)
100 76.66 64.84
50
0
silica
chitosan
PLGA
tissue culture slip
Fig. 7. The cell viability of the substrates with aligned patterned surface from silica colloids, chitosan, PLGA, The tests were performed with mMSCs and CKK-8 reagent, against a standard tissue culture cover slip.
mMSCs (Fig. 8). On the control glass substrate, the growth of mMSCs showed no evidence of alignment (Fig. 8A). On the PLGA patterned substrate, mMSCs formed a more complex arrangement, possibly due to the chemical composition of the substrate and/or the ladder-like aligned pattern of this substrate (Fig. 8B), whereas on the PLGA-chitosan and silica substrates, the cells displayed a spindle-shaped morphology and were aligned along the parallel ridge structures within the substrate (Fig. 8C and D). After culture for 5 days, confocal laser microscopy of phalloidin-stained cells confirmed the spindle shaped morphology and showed that the long axis of most cells was aligned parallel to the ridges within the PLGA-chitosan (Fig. S10) and silica substrates (Fig. S11). On the PLGA substrate, mMSCs appeared to reside in grooves between polymer ridges, and extended between comb-like gaps (Fig. S12).
397
Beyond their biocompatibility, the studies so far have demonstrated PLGA-chitosan and silica substrates are capable of directing the alignment of mMSC growth. The substrates can also be imaged by SEM at high magnification to visualize the location of mMSCs with the substrate. Due to the fact that the PLGA nanospheres can be easily detected under SEM, cells cultured for four days on PLGA-chitosan substrates were selected for SEM imaging. Figs. 9 and S13 show that the mMSCs are aligned parallel to ridged structures. The spindle-shaped cells are observed on the pattern ridges or in the region of the glass slide, but, in the latter case, the cells appear to extend filopodia to contact regions of polymer nanospheres. It has recently been shown that while lamellipodia mediate cell attachment and spreading on flat surfaces, filopodia play a key role in cell attachment to surfaces with nanotopographical features. Although the mechanisms have not been fully elucidated, it has been suggested that filopodia are unable to form stable adhesions on rigid, flat substrates, such as glass, and are rapidly replaced by lamellipodia, whereas on less rigid surfaces with nanotopographical features, the filopodia can form more stable adhesions, allowing the cell to apply the traction forces that are required for alignment [47,48]. Various scaffolds or substrates have been used to orientate the growth of stem cells [4–9,44–46]. With the newly prepared surface patterned substrates by the controlled freezing approach, this study with the mMSCs demonstrates the biocompatibility and the potential to direct cell growth. The limited issue for the current substrates is the stability of the patterned structures during cell culture. It had been observed that for some of the substrates the patterned structure could float off the glass cover slip in culture. This may be improved by varying the hydrophilicity and/or surface functionality of the glass surface or by use of other types of substrates (e.g., polystyrene, mica) to enhance the interaction of the patterned component and the substrate. Further study is required to understand the behaviour of mMSCs with the patterned
Fig. 8. Fluorescent microscopic images of the actin cytoskeleton of mMSCs on freeze-dried substrates following a 3 day culture period: plain glass (A), PLGA substrate (B), PLGA-chitosan substrate (C) and silica substrate (D). mMSCs were stained with phalloidin to visualize the actin cytoskeleton (green) and DAPI to visualize cell nuclei (blue). The actin cytoskeletons can be seen aligning with the patterned substrates. The same scale bars for all the images.
398
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399
evidence for the aligned growth of mMSCs on the PLGA-chitosan substrate. This study demonstrates that a versatile controlled freezing approach can be used to fabricate various patterned substrates that have the potential to be used for guiding stem cell growth. Acknowledgements The authors are grateful for the financial support by the EPSRC (EP/ F016883/1) and access to the facility in the Centre for Materials Discovery at the University of Liverpool. The authors also gratefully acknowledge the financial support from a project grant awarded by the BBSRC (BB/D014638/1). We thank Prof. S. Kazarian group of Imperial College London for FTIR imaging. Appendix A. Supplementary data Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.msec.2015.01.034. References
Fig. 9. SEM images of mMSCs attached and aligned to PLGA-chitosan substrate after culture for 4 days. The aligned mMSCs are parallel to ridges in a common direction, displaying a spindle-shaped morphology (A & B). In image B, the cells can be observed extending filopodia to contact regions of polymer (dashed line circles) whereas on glass, attachment is mediated by broader cell protrusions that resemble lamellipodia (solid line circle).
substrates and control of orientation angle of the aligned cells with the parameters including cell population & alignment angle, topology effect (e.g., groove width, ridge height, surface roughness), and substrate stiffness (e.g., by using of different nanoparticles and polymers). 4. Conclusions We have described a simple and versatile method to fabricate 2D surface patterned substrates. By spreading solutions or suspensions onto glass slides and directionally freezing using a computercontrolled freezing stage, followed by freeze drying, aligned surface patterns are readily produced. The preparation of aligned surface patterns for hydrophilic polymers, silica colloids, polymer/silica colloids, polymer colloids, and polymer/metal oxide nanoparticles, has been demonstrated. When the directional freezing procedures are applied twice in orthogonal directions, the substrates with grid surface patterns can be produced. The formation of such patterns is the result of orientated ice templating. The spacings between the aligned ridges may be tuned simply by varying freezing rates. In general, fast freezing rates lead to narrower spacings of the aligned patterns. The biocompatibility of the surface patterned substrates is demonstrated by WST-8 test and culturing with mMSCs. The chitosan pattern shows the best biocompatibility, very similar to the standard tissue culture slip. Compared to the plain glass slide where mMSC grow randomly, the aligned substrates (PLGA, PLGA nanospheres with chitosan, and silica colloids) can support and orientate the mMSC growth. While the stability and degradation of PLGA is an issue, the patterns fabricated with both PLGA-chitosan and silica colloid substrates are stable showing good performance in guiding mMSC alignment. The patterned substrate prepared from chitosan solution with suspended PLGA nanospheres gives better cell alignment and stability. SEM imaging provides further
[1] G.T. Daley, D.T. Scadden, Cell 132 (2008) 544–548. [2] P. Macchiarini, P. Jungebluth, T. Go, M.A. Asnaghi, L.E. Rees, T.A. Cogan, A. Dodson, J. Martorell, S. Bellini, P.P. Parnigotto, S.C. Dickinson, A.P. Hollander, S. Mantero, M.T. Conconi, M.A. Birchall, Lancet 372 (2008) 2023–2030. [3] M. Manassero, V. Viateau, M. Deschepper, K. Oudina, D. Logeart-Avramoglou, H. Petite, M. Bensidhoum, Tissue Eng. A 19 (2013) 1554–1563. [4] S.D. Subramony, B.R. Dargis, M. Castillo, E.U. Azeloglu, M.S. Tracey, A. Su, H.H. Lu, Biomaterials 34 (2013) 1942–1953. [5] M. Théry, J. Cell Sci. 123 (2010) 4201–4213. [6] M.P. Lutolf, P.M. Gilbert, H.M. Blau, Nature 462 (2009) 433–441. [7] C.Y. Tay, M. Pal, H. Yu, W.S. Leong, N.S. Tan, K.W. Ng, S. Venkatraman, F. Boey, D.T. Leong, L.P. Tan, Small 7 (2011) 1416–1421. [8] C. Mandoli, F. Pagliari, S. Pagliari, G. Forte, P.D. Nardo, S. Licoccia, E. Traversa, Adv. Funct. Mater. 20 (2010) 1617–1642. [9] X. Shi, S. Chen, J. Zhou, H. Yu, L. Li, H. Wu, Adv. Funct. Mater. 22 (2012) 3799–3807. [10] S.W. Hong, J. Huh, X. Gu, D.H. Lee, W.H. Jo, S. Park, T. Xu, T.P. Russell, PNAS 109 (2012) 1402–1406. [11] A.M. Hung, H. Noh, J.N. Cha, Nanoscale 2 (2010) 2530–2537. [12] C.M. Kolodziej, H.D. Maynard, Chem. Mater. 24 (2012) 774–780. [13] Y. Xu, F. Zhang, X. Feng, Small 7 (2011) 1338–1360. [14] H.C. Kim, S.M. Park, W.D. Hinsberg, Chem. Rev. 110 (2010) 146–177. [15] I.W. Hamley, Prog. Polym. Sci. 34 (2009) 1161–1210. [16] M.H. Stenzel, C. Barner-Kowollik, T.P. Davis, J. Polym. Sci. A Polym. Chem. 44 (2006) 2363–2375. [17] U.H.F. Bunz, Adv. Mater. 18 (2006) 973–989. [18] R.D. Piner, J. Zhu, F. Xu, S. Hong, C.A. Mirkin, Science 283 (1999) 661–663. [19] C.A. Mirkin, ACS Nano 1 (2007) 79–83. [20] J. Yamaguchi, K. Itaka, T. Hayakawa, K. Arai, M. Yamashiro, S. Yaginuma, H. Koinuma, Macromol. Rapid Commun. 25 (2004) 334–338. [21] W. He, C.R. Halberstadt, K.E. Gonsalves, Biomaterials 25 (2004) 2055–2063. [22] Z. Nie, E. Kumacheva, Nat. Mater. 7 (2008) 277–290. [23] P. Chandra, K. Lai, H.J. Sung, N. Murthy, J. Kohn, Biointerphases 5 (2010) 53–59. [24] Y. Xia, J.A. Rogers, K.E. Paul, G.M. Whitesides, Chem. Rev. 99 (1999) 1823–1848. [25] B.D. Gates, Q. Xu, J.C. Love, D.B. Wolfe, G.M. Whitesides, Annu. Rev. Mater. Res. 34 (2004) 339–372. [26] Y. Lu, S.C. Chen, Adv. Drug Deliv. Rev. 56 (2004) 1621–1633. [27] V.A. Schulte, Y. Hu, M. Diez, D. Bünger, M. Möller, M.C. Lensen, Biomaterials 31 (2010) 8583–8595. [28] S.A. Ruiz, C.S. Chen, Soft Matter 3 (2007) 168–177. [29] M. Darder, P. Aranda, M.L. Ferrer, M.C. Gutierrez, F. del Monte, E. Ruiz-Hitzky, Adv. Mater. 23 (2011) 5262–5267. [30] L. Qian, H. Zhang, J. Chem. Technol. Biotechnol. 86 (2011) 172–184. [31] S. Deville, Adv. Eng. Mater. 10 (2008) 155–169. [32] S.R. Caliari, B.A.C. Harley, Biomaterials 32 (2011) 5330–5340. [33] Q. Zhang, Y. Zhao, S. Yan, Y. Yang, H. Zhao, M. Li, S. Lu, D.L. Kaplan, Acta Biomater. 8 (2012) 2628–2638. [34] B.W. Riblett, N.L. Francis, M.A. Wheatley, U.G.K. Wegst, Adv. Funct. Mater. 22 (2012) 4920–4923. [35] H. Zhang, I. Hussain, M. Brust, M.F. Butler, S.P. Rannard, A.I. Cooper, Nat. Mater. 4 (2005) 787–793. [36] H. Zhang, J.Y. Lee, A. Ahmed, I. Hussain, A.I. Cooper, Angew. Chem. Int. Ed. 47 (2008) 4573–4576. [37] A. Ahmed, R. Clowes, E. Willneff, H. Ritchie, P. Myers, H. Zhang, Ind. Eng. Chem. Res. 49 (2010) 602–608. [38] A. Ahmed, W. Abdelmagid, H. Ritchie, P. Myers, H. Zhang, J. Chromatogr. A 1270 (2012) 194–203. [39] S. Deville, E. Maire, G. Bernard-Granger, A. Lasalle, A. Bogner, C. Gauthier, J. Leloup, C. Guizard, Nat. Mater. 8 (2009) 966–972. [40] L. Qian, H. Zhang, Green Chem. 12 (2010) 1207–1214.
L. Qian et al. / Materials Science and Engineering C 49 (2015) 390–399 [41] S.G. Kazarian, K.L.A. Chan, Analyst 138 (2013) 1940–1952. [42] S.G. Kazarian, K.L.A. Chan, V. Maquet, A.R. Boccaccini, Biomaterials 25 (2004) 3931–3938. [43] L. Glennon-Alty, R. Williams, S. Dixon, P. Murray, Acta Biomater. 9 (2013) 6041–6051. [44] M. Guvendiren, J.A. Burdick, Biomaterials 31 (2010) 6511–6518.
399
[45] S. Lenhert, M.B. Meier, U. Meyer, L. Chi, H.P. Wiesmann, Biomaterials 26 (2005) 563–570. [46] S. Fujita, M. Ohshima, H. Iwata, J. R. Soc. Interface 6 (2009) S269–S277. [47] J. Albushies, V. Vogel, Sci. Rep. 3 (2013) 1658. [48] B. Hoffmann, C. Schäfer, Cell Adhes. Migr. 4 (2010) 190–193.