Langmuir 2007, 23, 715-719
715
Spatially Controlled Cell Adhesion via Micropatterned Surface Modification of Poly(dimethylsiloxane) Natasha Patrito,† Claire McCague,† Peter R. Norton,*,† and Nils O. Petersen‡ Department of Chemistry, The UniVersity of Western Ontario, 1151 Richmond Street, London, ON, Canada, and National Institute for Nanotechnology, 11421 Saskatchewan DriVe, Edmonton, AB, Canada ReceiVed July 11, 2006. In Final Form: September 22, 2006 Spatial control of cell growth on surfaces can be achieved by the selective deposition of molecules that influence cell adhesion. The fabrication of such substrates often relies upon photolithography and requires complex surface chemistry to anchor adhesive and inhibitory molecules. The production of simple, cost-effective substrates for cell patterning would benefit numerous areas of bioanalytical research including tissue engineering and biosensor development. Poly(dimethylsiloxane) (PDMS) is routinely used as a biomedical implant material and as a substrate for microfluidic device fabrication; however, the low surface energy and hydrophobic nature of PDMS inhibits its bioactivity. We present a method for the surface modification of PDMS to promote localized cell adhesion and proliferation. Thin metal films are deposited onto PDMS through a physical mask in the presence of a gaseous plasma. This treatment generates topographical and chemical modifications of the polymer surface. Removal of the deposited metal exposes roughened PDMS regions enriched with hydrophilic oxygen-containing species. The morphology and chemical composition of the patterned substrates were assessed by optical and atomic force microscopies as well as X-ray photoelectron spectroscopy. We observed a direct correlation between the surface modification of PDMS and the micropatterned adhesion of fibroblast cells. This simple protocol generates inexpensive, single-component substrates capable of directing cell attachment and growth.
Introduction The spatial organization of cells plays a critical role in fundamental biological studies1 as well as in the development of tissue engineering scaffolds,2 biosensors,3,4 and microfluidic assays.5-7 To facilitate cell patterning, surfaces are functionalized with adhesion promoting and inhibiting molecules. Adhesive domains are often defined by islands of extracellular matrix proteins surrounded by poly(ethylene glycol)-based (PEG) surfactants and polymers that block indiscriminate cell attachment.8,9 Such substrates have traditionally been fabricated using photolithography10-12 and microcontact printing.13-16 These are time consuming and costly techniques which require multiple processing steps, complex chemistries, and a clean room facility. Although both methods routinely generate features with sub* Corresponding author. E-mail:
[email protected]. † The University of Western Ontario. ‡ National Institute for Nanotechnology. (1) Spargo, B. J.; Testoff, M. A.; Nelson, T. B.; Stenger, D. A.; Hickman, J. J.; Rudolf, A. S. Proc. Natl. Acad. Sci. 1994, 91, 11070-11074. (2) Langer, R.; Vacanti, J. P. Science 1993, 260, 920-926. (3) Morefield, S. I.; Keefer, E. W.; Chapman, K. D.; Gross, G. W. Biosens. Bioelectron. 2000, 15, 383-396. (4) Park, T. H.; Shuler, M. L. Biotechnol. Prog. 2003, 19, 243-253. (5) Chiu, D. T.; Jeon, N. L.; Huang, S.; Kane, R. S.; Wargo, C. J.; Choi, I. S.; Ingber, D. E.; Hickman, J. J. Proc. Natl. Acad. Sci. 2000, 97, 2408-2413. (6) Rhee, S. W.; Taylor, A. M.; Tu, C. H.; Cribbs, D. H.; Cotman, C. W.; Jeon, N. L. Lab-on-a-Chip 2005, 5, 102-107. (7) Tan, W.; Desai, T. A. Tissue Eng. 2003, 9, 255-267. (8) Raghavan, S.; Chen, C. S. AdV. Mater. 2004, 16 (15), 1303-1313. (9) Liu, W. F.; Chen, C. S. Mater. Today 2005, 8 (12), 28-35. (10) Folch, A.; Toner, M. Biotechnol. Prog. 1998, 14, 388-392. (11) Rozkiewicz, D.; Kraan, Y.; Werten, M. W. T.; de Wolf, F. A.; Subramanian, V.; Ravoo, B. J.; Reinhoudt, D. N. Chem. Eur. J. 2006, 12 (19), 6290-6297 (12) Yi, D. K.; Kim, D. Y.; Turner, L.; Breuer, K. S.; Kim, D. Y. Biotechnol. Lett. 2006, 28, 169-173. (13) De Silva, M. N.; Desai, R.; Odde, D. J. Biomed. MicrodeVices 2004, 6, 219-222. (14) De Silva, M. N.; Paulsen, J.; Renn, M. J.; Odde, D. J. Biotechnol. Bioeng. 2006, 93, 919-927. (15) Jiang, X.; Takayama, S.; Qian, X.; Ostuni, E.; Wu, H.; Bowden, N.; LeDuc, P.; Ingber, D. E.; Whitesides, G. M. Langmuir 2002, 18, 3273-3280. (16) Chen, C. S.; Tan, J.; Tien, J. Annu. ReV. Biomed. Eng. 2004, 6, 275-302.
micrometer resolution, the monolayer modifications they yield are often too fragile to withstand physiological or microfluidic shear stresses.6 Due to these limitations, there remains great interest in the development of simple and inexpensive techniques for rapid and reproducible cell patterning. Poly(dimethylsiloxane) (PDMS) is a useful substrate for the fabrication of cell arrays because it is biocompatible, inexpensive, and durable and can be readily integrated with microsystems.17 Due to the material’s low surface energy and inherent hydrophobicity, PDMS surfaces inhibit cell adhesion. Consequently, the surface chemistry of PDMS must be tailored to improve its bioactivity and enable patterned cell proliferation. De Silva and co-workers recently achieved localized cell attachment by patterning adhesive proteins on PDMS via microcontact printing13 and precision aerosol spraying.14 These techniques exploited the inherent hydrophobicity of PDMS and did not require the additional patterning of adhesion-inhibiting molecules, such as PEG or Pluronic surfactants. Compared to traditional photolithography-based methods, these protocols were rapid and simple; however, they continued to rely upon the patterned deposition of extracellular matrix proteins such as laminin and collagen. Herein we report the fabrication and characterization of a robust PDMS substrate capable of spatially confining cell proliferation without the use of mediating proteins or molecules. Chemical activation of the polymer surface is achieved in a single step by sputter deposition of a thin-metal film through a stencil-mask, in the presence of a gaseous plasma. Removal of the metal layer exposes regions of the polymer surface which are enriched in oxygen, rendering them less hydrophobic and more bioactive. Fibroblast cells seeded on the PDMS substrates selectively adhered to these modified areas and formed multilayer structures when cultured beyond confluency. In addition, we show that the patterned PDMS substrates retain their biological activity for at least 25 days if stored with the metal film intact. Owing to its (17) Mata, A.; Fleischman, A. J.; Roy, S. Biomed. MicrodeVices 2005, 7, 281-293.
10.1021/la062007l CCC: $37.00 © 2007 American Chemical Society Published on Web 12/15/2006
716 Langmuir, Vol. 23, No. 2, 2007
Patrito et al.
Figure 1. Optical micrographs of patterned thin film PDMS sample (a, b) after aluminum deposition, (c, d) after etching of aluminum, (e, f) with localized water droplets, (g, h) after COS-7 cell culture.
simplicity, the described technique is amenable to low-cost mass production and rapid prototyping of cell patterns. Materials and Methods PDMS Substrate Preparation. PDMS precursors, Sylgard 184A and B (Dow Corning), were combined in a 10:1 mass ratio. Bulk films were prepared by pouring the prepolymer mixture into a glass Petri dish to a thickness of 5 mm. Thin film PDMS samples were generated by spin casting a 5% (w/w) solution of PDMS precursors in heptane onto clean silicon wafers at 3000 rpm for 30 s with a Solitec 850 spin coater (Solitec Wafer Processing). Both the bulk and thin film PDMS samples were cured at 70 °C for 90 min. The thicknesses of the thin film samples were measured to be 220 ( 3 nm using a Gaertner Scientific L2W16D 1.3 ellipsometer (Supporting Information). PDMS Surface Modification. Microdots of Al were deposited on thin film and bulk PDMS samples in an identical fashion. Stainless steel screens (InterNet Inc.) with etched holes of 180 and 230 µm diameters were employed as physical masks. Al was sputtered onto the PDMS substrates through the steel screens using an Edwards Auto 500 Magnetron Sputtering System. The sputter system was operated at 500 W, 5.5 mTorr, with Ar gas flowing at a rate of 15 sccm. Prior to metal deposition, the samples were kept in the chamber with the plasma struck and the source shutter closed for 10 min. The thickness of the deposited aluminum layer was determined to be 42 ( 4 nm using a Dektak3 profilometer (Veeco). The deposited aluminum was subsequently etched by immersion in 1.8 M orthophosphoric acid for 20-30 min to reveal the underlying PDMS which was both chemically and topographically modified. Surface Analysis. Optical images of the PDMS samples were captured after each processing step using a Zeiss Axioskop2 Mat Microscope with a QImaging Retiga 1300 CCD digital camera. Static water contact angle measurements were made using a Model 100 Manual Goniometer (Rame-Hart) to assess changes in the hydrophobic character of the micropatterned PDMS. A droplet of deionized water was placed on both pristine and treated polymers at room temperature, and after 10 s a contact angle was recorded. Contact angle data collected from five pristine PDMS and five modified PDMS samples were averaged. To assess topographical changes in the treated silicone, atomic force microscopy (AFM) images were collected on a Nanoscope IIIa (Digital Instruments) in contact mode using NP-S20 cantilevers (k ) 0.06 N/m) from Veeco Instruments Inc. The elemental composition of the thin film PDMS samples was analyzed by X-ray photoelectron spectroscopy (XPS) using a Kratos Axis Ultra spectrometer (Kratos Analytical). The XPS spectrometer employed an Al KR X-ray source (1486.6 eV) operated at a source energy of 210 W and a charge neutralizer system
with a filament current of 1.6 A and a charge balance of 2.4 V. XPS survey and high-resolution spectra of the C1s, O1s, and Si2p binding energy regions were collected at pass energies of 160 and 20 eV, respectively. XPS spectra were fit and signal areas calculated using CasaXPS software. The binding energy scale was calibrated using the C1s peak at 284.8 eV. Bioactivity Assessment. COS-7 fibroblast cells (ATCC) were cultured on the PDMS substrates. Dulbecco’s Modified Eagle Medium (DMEM, Invitrogen) with 2 mM L-glutamine, 1% penicillinstreptomycin (Invitrogen), and 10% fetal bovine serum (FBS, Invitrogen) was used for all experiments. Cell cultures were incubated at 37 °C and 100% humidity. Samples of surface modified PDMS (1 cm2) were sterilized by rinsing with 95% ethanol. Cells (7.5 × 103/cm2) were seeded onto the polymer substrates and cultured for 36 h prior to fixation by immersion in an aqueous 4.4% paraformaldehyde (Sigma Aldrich) solution.
Results and Discussion Patterns of aluminum were deposited on PDMS in a square array of 180 µm diameter dots with center-to-center spacing of 250 µm and a hexagonal array of 230 µm diameter dots with center-to-center spacing of 395 µm (Figure 1a,b). The aluminum was etched from the PDMS surface, and the exposed areas were indistinguishable by optical microscopy from the surrounding pristine polymer. XPS analyses of the etched polymer surfaces confirm the removal of the aluminum coating. After etching, aluminum comprises less than 0.7% of the atomic composition of the modified PDMS surface, and no phosphoric acid byproducts were observed above the detection limit (Supporting Information). The randomly distributed features which appear as dark speckles in Figure 1a-h are artifacts of the spin coating process and were not present in the bulk films. These features do not impact the further processing of PDMS. The etched samples were submerged in distilled deionized water and were slowly withdrawn. Square and hexagonal arrays of water droplets were formed on the surface, revealing an increase in the hydrophilicity of the treated areas. COS-7 cells were cultured on the modified PDMS samples for 36 h and were subsequently fixed. Optical micrographs showed the localized proliferation of cells on the modified regions (Figure 1g,h). Figure 2 shows the correlation between the size of the deposited aluminum dots and the areas occupied by the fibroblast cells. XPS was employed to probe chemical changes within the uppermost 10 nm of the treated PDMS substrates, and the results
Micropatterned Cell Adhesion on Surface Modified PDMS
Figure 2. Optical micrographs of thin film PDMS samples showing aluminum dots of diameter (a) 230 µm and (b) 180 µm. COS-7 cells localized on treated regions of diameter (c) 230 µm and (d) 180 µm.
Langmuir, Vol. 23, No. 2, 2007 717
Figure 3. High-resolution Si 2p XPS spectra of pristine and modified PDMS.
Table 1. XPS Atomic Composition Analyses of Native and Modified PDMS binding energy (eV)
atomic composition (%)
C 1s O 1s Si 2p
Native PDMS 284.8 532.6 102.1
51.5 28.8 19.7
C 1s O 1s Si 2p
Surface-Modified PDMS 284.8 532.6 102.8
36.9 42.5 19.9
are shown in Table 1 and Figure 3. Atomic composition analyses indicate that, while the silicon content remains essentially unchanged, the oxygen content of the PDMS surface increases while its carbon content decreases. These results are comparable to those obtained for the treatment of PDMS with oxidizing plasmas, including argon.18-21 It is proposed that high energy species introduced by the plasma impinge upon the polymer surface, breaking silicon-carbon bonds and generating lowmolecular mass PDMS chains through scission.19,22 Carbon leaves the surface in the form of volatile species, and oxidative crosslinking via the formation of Si-O bridges generates a silica-like film.22 This is evidenced by the observed shift of the Si2p photoelectrons toward higher binding energies, indicative of the formation of an inorganic silica-like surface layer.18 Contact angles of 104 ( 2° and 30 ( 1° were measured for the pristine and modified PDMS, respectively. The increased hydrophilicity of the modified PDMS is consistent with the increased oxygen content at the polymer surface observed by XPS. The post-aluminum etch topography of thin film and bulk PDMS samples were examined by AFM. Random sinusoidal ripples (wavelength ∼ 1 µm, amplitude ∼ 30 nm) were observed in the areas exposed to the aluminum deposition process. As shown in Figure 4a, the rippling dissipates around the perimeter (18) Hillborg, H.; Ankner, J. F.; Gedde, U. W.; Smith, G. D.; Yasuda, H. K.; Wikstro¨m, K. Polymer 2000, 41, 6851-6863. (19) Hillborg, H.; Tomczak, N.; Ola`h, A.; Scho¨nherr, H.; Vancso, G. J. Langmuir 2004, 20, 785-795. (20) Vladkova, T. G.; Keranov, I. L.; Dineff, P. D.; Youroukov, S. Y.; Avramova, I. A.; Krasteva, N.; Altankov, G. P. Nucl. Instrum. Methods Phys. Res. B 2005, 236, 552-562. (21) Katzenberg, F. Surf. Coat. Technol. 2005, 200, 1097-1100. (22) Kim, J.; Chaudhury M. K.; Owen, M. J. J. Colloid Interface Sci. 2000, 226, 231-236.
Figure 4. Contact mode AFM height images of (a) the boundary between modified and unmodified regions of a thin film PDMS sample and (b) a higher resolution image of the observed ripples.
of the microdot in a band less than 5 µm wide. The root-meansquared roughness of the modified PDMS was 17.3 nm as compared to 1.6 nm for the surrounding unmodified surface. Similar rippling was observed on the bulk PDMS samples. The ripples are formed to relieve thermal stresses incurred during the deposition of aluminum. These stresses arise from differences in the thermal expansion coefficients of the metal film, the silicalike material generated, and the underlying PDMS.24-29
718 Langmuir, Vol. 23, No. 2, 2007
Patrito et al.
Figure 6. Optical micrographs of (a) patterned bulk PDMS aluminum etched after 25 days of storage in air, (b) water droplets localized on the modified surface, and (c) stored sample cultured with fibroblast cells for 24 h. Figure 5. (a) Contact mode AFM height image of molded bulk PDMS surface. Optical micrographs of molded PDMS (b) before and (c) after plating of COS-7 fibroblast cells.
The process of aluminum deposition thus results in both chemical and topographic modifications to the polymer surface. To resolve the relative impact of these changes on cell proliferation, equivalent ripple patterns were prepared in PDMS via soft lithography. PDMS substrates patterned with microdots of aluminum were used as molds onto which PDMS precursors were poured to a thickness of 5 mm. After curing for 90 min at 70 °C, the PDMS was peeled from the mold and AFM imaging confirmed the transfer of the ripple pattern (Figure 5a). Patterned wetting of the surface by deionized water was not observed, demonstrating that the molded topography did not increase the hydrophilicity of the PDMS (Figure 5b). After 36 h cell culture, COS-7 cells had not proliferated on the molded PDMS substrates (Figure 5c). This result indicates that the oxygen enrichment incurred at the PDMS surface during metal deposition is the primary basis for the observed improvement in bioactivity. While the topographic changes to the PDMS surface are permanent, further experiments were required to evaluate the stability of the chemical modifications. Previous studies have shown that plasma modified bulk PDMS demonstrates a substantial hydrophobic recovery within hours of treatment.22,30-31 Hydrophobic recovery is mediated by the migration of lowmolecular weight PDMS chains from the bulk of the polymer to its surface through cracks or pores in the hydrophilic silicalike layer.22,32 This recovery phenomenon significantly limits the lifespan of the activated PDMS surface and restricts the commercial application of plasma-modified PDMS substrates. We therefore evaluated the long-term stability of the surfaces produced. The aluminum-coated bulk PDMS samples were stored in air for 25 days. The aluminum was then etched, and the surface properties of the stored substrates were assessed. As shown in Figure 6, localized water droplets were observed and fibroblast (23) Ouyang, M.; Yuan, C.; Muisener, R. J.; Boulares, A.; Koberstein, J. T. Chem. Mater. 2000, 12, 1591-1596. (24) Bowden, N.; Brittian, S.; Evans, A. G.; Hutchinson, J. W., Whitesides, G. M. Nature (London) 1998, 393, 146-149. (25) Bowden, N.; Huck, W. T. S.; Paul, K. E.; Whitesides, G. M. Appl. Phys. Lett. 1999, 75, 2557-2559. (26) Tserepi, A.; Gogolides, E.; Tsougeni, K.; Constantoudis, V.; Valamontes, E. S. J. Appl. Phys. 2005, 98, 113502. (27) Chua, D. B. H.; Ng, H. T.; Li, S. F. Y. Appl. Phys. Lett. 2000, 76, 721723. (28) Zhao, S.; Denes, F.; Manolache, S.; Carpick, R. W. Proceedings of the SEM VII International Congress and Exposition on Experiment and Applied Mechanics, 2002, pp 162-165. (29) Huck, W. T. S.; Bowden, N.; Onck, P.; Pardoen, T.; Hutchinson, J. W.; Whitesides, G. M. Langmuir 2000, 16, 3497-3501. (30) Klemic, K. G.; Klemic, J. F.; Reed M. A.; Sigworth, F. J. Biosens. Bioelectron. 2002, 17, 597-604. (31) Makamba, H.; Kim, J. H.; Lim, K.; Park, N.; Hahn, J. H. Electrophoresis 2003, 24, 3607-3619. (32) Hillborg H.; Gedde U. W. Polymer 1998, 39, 1991-1998.
Figure 7. Stored sample cultured with fibroblast cells for 36 h.
cells readily proliferated on the micropatterned surface. Cells remained patterned on the substrates over a 72 h observation period in the presence of serum-containing media, indicating that the modified polymer substrates resist fouling. These results suggest that the aluminum acted as a protective layer, maintaining the enhanced hydrophilicity and bioactivity of the modified PDMS regions throughout extended storage. Within 24 h of exposure to air, the etched surface of the bulk sample regained its hydrophobic nature (Supporting Information). As shown in Figure 7, cells cultured beyond confluency on the stored substrates did not escape the modified PDMS regions and, instead, proliferated in multiple layers. Elongated membrane protrusions (>150 µm) were observed spanning the gaps between patterned cell colonies. Similar bridging projections were reported by Rozkiewicz and co-workers after the incubation of HeLa cells on silicon covalently micropatterned with 100 µm diameter protein islands.11 The presence of these membranous extensions may suggest a means of cell-to-cell communication between fibroblasts in neighboring dot colonies and merits further investigation.
Conclusions Patterned magnetron-sputter deposition of aluminum onto PDMS is a simple, rapid, and cost-effective means of preparing addressable cell arrays. Through this process, localized topographical and chemical modifications which enhance the bioactivity of the polymer surface were simultaneously achieved. COS-7 fibroblast cells adhered preferentially to the treated PDMS, proliferating in confined microdot domains. Topographic patterning of the polymer substrate without chemical modification was also achieved via soft lithography. In the absence of the surface-oxygen enrichment generated by plasma exposure during aluminum deposition, cells did not adhere to the roughened surfaces. The chemical modification of PDMS was thus shown to be the primary director of the observed spatially controlled cell proliferation. Throughout an extended period of storage, the deposited aluminum acted as a barrier to the hydrophobic recovery of PDMS, protecting the enhanced wettability and bioactivity of the micropatterned polymer. This method generates single-component platforms for cell arrays, requiring no photolithography or preadsorption of adhesive
Micropatterned Cell Adhesion on Surface Modified PDMS
molecules. An additional advantage of this technique is that the modified PDMS substrates resist fouling and cell patterns are maintained in the presence of serum. Through the use of flexible or shape-matched physical masks, the described technique may also be amenable to the patterning of curved surfaces. The aluminum coating ensures that the chemically modified substrates can be conveniently stored in air with little or no performance loss. The simplicity of the described technique makes it amenable to low-cost mass production and rapid prototyping of cell patterns. Such arrays offer means to probe cell-cell interactions, cell motility, and cell signaling among other bioanalytical concerns by accessing cellular responses to varied spatial or geometric organization. Future research will focus on creating arrays of multiple cell types, generating single-progenitor cell colonies,
Langmuir, Vol. 23, No. 2, 2007 719
and patterning other biologically relevant molecules including proteins. Acknowledgment. This research was supported by the Natural Sciences and Engineering Research Council of Canada and the Canadian Institute for Photonics Innovation. We thank Mr. M. Biesinger and Mr. B. Kobe for XPS technical assistance and Mr. Phil Shaw of the Physics machine shop. Supporting Information Available: Thicknesses of films measured via ellipsometry, PDMS spectrum, and static water contact angle of aged PDMS samples. This material is available free of charge via the Internet at http://pubs.acs.org. LA062007L