Biomedical Microdevices

, Volume 12, Issue 1, pp 13–21

Targeted cell adhesion on selectively micropatterned polymer arrays on a poly(dimethylsiloxane) surface


  • Linzhi Tang
    • Gachon BioNano Research Institute & Division of BioNano Technology and College of BioNano TechnologyKyungwon University
  • Junhong Min
    • Gachon BioNano Research Institute & Division of BioNano Technology and College of BioNano TechnologyKyungwon University
  • Eun-Cheol Lee
    • Gachon BioNano Research Institute & Division of BioNano Technology and College of BioNano TechnologyKyungwon University
  • Jong Sung Kim
    • Department of Chemical & BioengineeringKyungwon University
    • Gachon BioNano Research Institute & Division of BioNano Technology and College of BioNano TechnologyKyungwon University

DOI: 10.1007/s10544-009-9353-1

Cite this article as:
Tang, L., Min, J., Lee, E. et al. Biomed Microdevices (2010) 12: 13. doi:10.1007/s10544-009-9353-1


Herein, we introduce the fabrication of polymer micropattern arrays on a chemically inert poly(dimethylsiloxane) (PDMS) surface and employ them for the selective adhesion of cells. To fabricate the micropattern arrays, a mercapto-ester—based photocurable adhesive was coated onto a mercaptosilane—coated PDMS surface and photopolymerized using a photomask to obtain patterned arrays at the microscale level. Robust polymer patterns, 380 µm in diameter, were successfully fabricated onto a PDMS surface, and cells were selectively targeted toward the patterned regions. Next, the performance of the cell adhesion was observed by anchoring cell adhesive linker, an RGD oligopeptide, on the surface of the mercapto-ester—based adhesive-cured layer. The successful anchoring of the RGD linker was confirmed through various surface characterizations such as water contact angle measurement, XPS analysis, FT-IR analysis, and AFM measurement. The micropatterning of a photocurable adhesive onto a PDMS surface can provide high structural rigidity, a highly–adhesive surface, and a physical pathway for selective cell adhesion, while the incorporated polymer micropattern arrays inside a PDMS microfluidic device can serve as a microfluidic platform for disease diagnoses and high-throughput drug screening.


Polymer micropattern arrayMercapto-ester adhesiveMercaptosilanePhotopolymerizationCell adhesionRGD linker

1 Introduction

The attachment of cells onto a substrate has been one of the main issues for the high-throughput screening (HTS) of drugs, cancer studies, disease diagnoses, and fundamental biological researches, and is the primary requirement for obtaining information on cellular interactions (Barron et al. 2004). With the growing interest in Lab-on-a-chip technologies concerned with cell research, the robust attachment of cells inside a microfluidic system or in the form of a cellular microarray on a chemically-inert poly(dimethylsiloxane) (PDMS) surface is of great significance, and the targeted patterning of cells inside microfluidic devices is becoming a primary concern. However, it is not easy to systematically align cells directly inside a PDMS microfluidic device.

Geometrical control (Goto et al. 2008; Jiang et al. 2002; Mata et al. 2002) of the attachment and growth of cells is one direction for in vitro cell studies. The Kitamori’s group (Goto et al. 2008) fabricated micro- and nano-patterned surfaces using controlled surface functionality in order to investigate the interaction between cells and the extracellular matrix (ECM) inside a microchannel. Chemical modification (Fukuda et al. 2006; Ito, 1999; Itoga et al. 2006; Lehnert et al. 2004; Lovchik et al. 2008; Ochsner et al. 2007; Richert et al. 2004; Svedhem et al. 2003) of the surface is another direction. The Whitesides’ group (Lee et al. 2004) studied the influence of PDMS surface composition on cell growth by varying the PDMS prepolymer-to-curing agent ratio, as well as the oxidation state. Many researchers have endeavored to align the cells on various surface types adopting ECM proteins such as fibronectin, collagen, elastin, and laminin (Cimetta et al. 2009; Goto et al. 2008; Jiang et al. 2002; Lo et al. 2000; Mata et al. 2002). However, these are highly expensive, and are difficult to selectively pattern without the aid of an elastomeric stamp or geometrical barriers.

Herein, we propose a selective cell patterning strategy by adopting a mercapto-ester—based photocurable adhesive, a Norland Optical Adhesive (NOA), as an intermediate layer for the targeted adhesion of cells on a chemically-inert PDMS surface for first step toward establishing incorporated cell arrays applicable for microfluidic experiment. Cells generally grow well on ECM-coated surfaces, but it has also been reported that cell movement and attachment are influenced by the rigidity of the substrate (Gray et al. 2003; Lo et al. 2000; Tzvetkova-Chevolleau et al. 2008). Mercapto-ester—based photocurable adhesives exhibit a high mechanical property upon photopolymerization. Due to their photocurable property, high optical transparency, and high bonding performance with various hard and soft substrates, mercapto-ester—based adhesives have been generally adopted as optical glues for MEMS device fabrication (Cadarso et al. 2008) and as a medical adhesive for gluing tissues (Calaghan et al. 2000). In our previous studies (Lee et al. 2006a; Lee et al. 2006b), NOA was used for the fabrication of micro-well structures for the selective confinement of antibodies, and as a platform for the immobilization of Escherichia coli for the construction of biosensors. Mercapto-ester—based photocurable adhesives can be utilized as an intermediate matrix for targeted and robust patterning on an inert PDMS surface due to their photocurable nature, which is advantageous for selective patterning, as well as their high structural rigidity. The photocurable property of the mercapto-ester—based adhesives facilitate the formation of cell arrays in a simple and facile manner, and its high mechanical strength offers a robust platform for cell adhesion.

In this study, we demonstrate a micropattern formation employing a mercapto-ester—based photocurable adhesive, NOA, to provide physical pathways for establishing a cell array platform on a chemically-inert PDMS surface, and we apply the micropatterned surface toward the targeted and robust adhesion of cells. In addition, we demonstrate the effect of RGD oligopeptide (Burdick et al. 2004; Yamauchi et al. 2003), a well-known cell adhesive linker, on the adhesion of cells by anchoring the RGD linker onto the surface of the NOA. Various surface characterizations such as contact angle, XPS, FT-IR, and AFM analyses were made on both bare and RGD-anchored NOA surfaces to verify the effect of the RGD linker on the adhesion performance of cells.

2 Materials and methods

2.1 Materials

Commercialized mercapto-ester—based photocurable adhesives, Norland Optical Adhesives (NOA) 61, 63, and 71, were purchased from Norland Company (Cranbury, NJ, USA). Poly(dimethylsiloxane) (PDMS) prepolymer (Sylgard 184) and a curing agent were purchased from Dow Corning (Midland, MI, USA). Arginine-glycine-aspartic acid (RGD) oligopeptide linker (KRGDRGDRGD) was synthesized (> 95%) from Peptron (Korea). Phosphate buffered saline (PBS) buffer solution (pH 7.4) was purchased from Invitrogen, and poly(ethylene terephthalate) (PET) film was obtained from SKC (Korea). 3-mercaptopropyltrimethoxysilane (mercaptosilane) was purchased from Gelest (Morrisville, PA, USA). Human Mammary Epithelial Cells (HMEC) were purchased from Lonza (Wakersville, MD, USA). RPMI 1640 culture medium, fetal bovine serum (FBS), penicillin-streptomycin, and GlutaMAX were purchased from Gibco (Grand Island, NY, USA). Live-dead cell staining kit was purchased from BioVision (Mountain View, CA, USA).

2.2 Surface characterizations

2.2.1 Contact angle measurement

The water contact angles were measured on both bare and RGD-anchored NOA surfaces by the sessile drop (5 μL) technique using a KRÜSS DSA10–MK2 contact angle measuring system (KRÜSS GmbH, Germany). The reaction conditions for the anchoring of the RGD linker on NOA surface were 3 h and 16 h at room temperature (RT) and 37°C. The measurements were analyzed with Drop Shape Analysis software. Three measurements were made and averaged.

2.2.2 XPS analysis

X-ray photoelectron spectroscopy (XPS) analyses were conducted using a PHY 5700 (PHI, Chanhassen, MN, USA) equipped with an aluminum X-ray radiation source (1486.6 eV) and pass energy of 23.5 eV. The pressure in the chamber was below 1.3 × 10−9 Torr before the data were taken, and the voltage and current of the anode were 15 kV and 26.7 mA, respectively. The take-off angle was set at 45o. The binding energy of Au 4f7/2 (84.0 eV) was used as the reference. The resolution for the measurement of the binding energy was about  ± 0.6 eV. XPS analyses were interpreted using an ESCA1 (PHI, Chanhassen, MN, USA).

2.2.3 FT-IR spectroscopy

The presence of amino acids peaks were measured using a Fourier Transform Infrared (FT-IR) Spectrometer (Vertex 70, Bruker) in attenuated total reflection (ATR) mode, equipped with a deuterated triglycine sulfate (DTGS) detector.

2.2.4 AFM analysis

Atomic force microscopy (AFM) was performed in non-contact mode using a PSIA XE-150 (PSIA Inc.), with a 0.5 Hz scan rate in atmosphere at RT. A non-contact ultrasharp silicon cantilever (NSC 15) was used, and the scan area for surface roughness was 10 μm × 10 μm.

2.3 Cell culture

Cells were cultured using RPMI-1640 culture medium containing 10% FBS, 1% penicillin-streptomycin, and 1% GlutaMAX, and maintained in a humidified 5% CO2 / 95% air atmosphere at 37°C with 5 × 105 / well of initial seeding amount. After cultivation, cells were rinsed thoroughly to remove cells not adhered on the surface. Live and dead cells were distinguished by dying the immobilized cells using the staining kit containing Carcein green dye (Excitation: 488 nm / emission: 518 nm) and PI red dye (Excitation: 488 nm / emission: 615 nm) for the detection of both live and dead cells, respectively. All microscopic experiments were performed using cells cultured for 1 day and additionally cultured on the patterned surface for 1 day. Optical and fluorescence images were taken using a fluorescence microscope (Nikon Eclipse TE 2000-U) and were analyzed using NIS-Elements software.

2.4 Cell adhesion on NOA-coated PET surface

The NOA adhesive was coated onto a poly(ethylene terephthalate) (PET) substrate and photopolymerized (λ = 365 nm, 45 mW/cm2) overnight. Cells were grown overnight (24 h) on the NOA surface, and the adhesion and growth were detected using an optical microscope. 1.8 mM RGD linker was diluted in a PBS buffer (pH 7.4). Anchoring of the RGD linker was performed for 3 h and 16 h at both RT and 37°C. Cell growth was observed on the RGD-anchored NOA surface, and was detected using an optical microscope. Live and dead cells were stained using a live-dead cell staining kit, and were detected using a fluorescence microscope.

2.5 NOA micropatterning on PDMS surface

Micropatterning of the NOA was performed on a PDMS surface as shown in Fig. 1. To facilitate the adhesion of the NOA on the PDMS surface, the PDMS surface was first treated with O2 plasma (170 W) for 1 min (Fig. 1(a)), immersed in a 1% (v/v) aqueous solution of 3-mercaptopropyltrimethoxysilane (mercaptosilane) for 30 min while heating at 80°C, and washed thoroughly with distilled water and then dried (Fig. 1(b)). Next, the NOA was spin-coated onto the mercaptosilane-coated PDMS at 3,000 rpm for 30 s to obtain a homogeneous thickness (Fig. 1(c)). The spin-coated NOA was then photopolymerized (λ = 365 nm, 18 W/cm2) using a UV power source (Exfo OmniCure 1000) (Fig. 1(d)). The distance between the NOA-coated PDMS and UV outlet was set to 5 cm. A photomask having circular patterned arrays, 300 μm in diameter, was used for selective micropatterning. Unpolymerized NOA was removed by immersing the micropatterned PDMS in isopropyl alcohol and sonicating it for 12 ∼ min (Fig. 1(e)). The micropatterned NOA was post-cured under a UV light for 2 h for a robust pattern formation. Cell adhesion and growth on the selectively micropatterned NOA layer atop the PDMS was detected using an optical and a fluorescence microscopes.
Fig. 1

Schematic flow for the fabrication of micropatterned arrays of mercapto-ester—based photocurable adhesive on a PDMS surface. (a) PDMS surface oxidation by O2 plasma. (b) PDMS surface silanization using a mercaptosilane. (c) NOA coating on the mercaptosilane-modified PDMS. (d) UV (λ = 365 nm, 18 W/cm2) illumination. (e) Washing off of the unpolymerized NOA using isopropyl alcohol

3 Results and discussion

3.1 Selection of optimum NOA surface for cell adhesion

The NOA was selected as an intermediate layer for cell adhesion on the PDMS surface because of its selective patterning performance due to its photocurable property and structural rigidity. These are the first criteria to select an optimum intermediate layer that bonds strongly to various substrate types for the successful targeting of cells on a substrate. The adhesion of all three types of NOA on PET was robust. However, the adhesions of NOA 63 and NOA 73 on the glass substrates were relatively weak. NOA patterns made of 63 and 73 were easily peeled off while drying the patterned substrates with an air gun after photopolymerization on the glass substrates. The adhesion of NOA 61 on a glass substrate was robust. Since the target substrate for cell adhesion in this study is PDMS, and oxidized PDMS resembles a glass surface, NOA 61 was chosen as the optimum intermediate layer for cell growth. NOA 61 is a mercapto-ester—based photocurable adhesive with a relatively low viscosity, 300 cps. Because of its relatively low viscosity, it is easy to spin-coat and micropattern. In addition, its modulus of elasticity and tensile strength are approximately 1034 MPa and 20.7 MPa, respectively, which are more than 600 times and 3.3 times higher than those of PDMS (1.5 MPa and 6.2 MPa), respectively. Taking these into consideration, NOA 61 is rendered appropriate for fabricating robust patterns with a relatively easy fabrication process.

3.2 Surface characterizations

3.2.1 Contact angle measurement

The RGD linker was anchored onto the surface of the NOA 61-coated PET, and the cell adhesion performance was compared with that of a bare NOA 61 surface. Four reaction conditions — 3 h and 16 h anchoring at both RT and 37°C — were tested for the anchoring of the RGD linker on the NOA 61 surface. The water contact angles were measured on bare and RGD-anchored NOA 61 surfaces, and the results of the water contact angles measured after anchoring for 3 h at RT and 37°C were shown in Fig. 2. As can be seen, the water contact angle decreased from 88.6o to 78.8o and 82.2o after anchoring with the RGD linker at RT and 37°C, respectively. The RGD linker is mainly composed of arginine, glycine, and aspartic acid, and the arginine and aspartic acid contain polar side groups. This might have decreased the water contact angle on the RGD-anchored NOA 61 surface. The water contact angle difference seemed negligible for the RT and 37°C anchoring conditions. Extending the anchoring time to 16 h did not greatly enhance the anchoring performance (data not shown).
Fig. 2

Contact angle measurements on RGD-anchored NOA 61 layer on PET. (a) Bare NOA 61. (b) RGD anchored at RT for 3 h. (c) RGD anchored at 37°C for 3 h

3.2.2 XPS and FT-IR analyses

To confirm the successful anchoring of the RGD linker on the NOA 61 surface, XPS analyses were performed. As shown in Fig. 3(b) and (c), an N1s peak appeared as a result of the 3 h RGD anchoring because lysine and arginine contain nitrogen in their side groups, and amino acid itself contains an amine terminal. The peak intensity of N1s, however, was slightly higher in Fig. 3(b) than in (c), probably due to the lesser anchoring of the RGD linker, which coincides well with the results of the water contact angle measurements in Fig. 2. The carbon and oxygen contents were kept relatively unchanged, but the Si peaks decreased significantly after the RGD anchoring. The anchoring of the RGD linker on the coated NOA 61 layer might have accompanied the reaction with the Si groups existing on the NOA 61 surface. From the structural perspective, NOA is mainly composed of two parts — acrylate group and mercapto group (Pinto-Iguanero et al. 2002). Besides the interaction between the RGD and Si group of NOA 61, the amine residues on lysine on RGD might have reacted with the acrylate and mercapto groups on NOA 61 for the anchoring of the RGD linker on micropatterned NOA 61.
Fig. 3

XPS analyses on RGD-anchored NOA 61 layer on PET. (a) Bare NOA 61. (b) RGD anchored at RT for 3 h. (c) RGD anchored at 37°C for 3 h

To further confirm the successful anchoring of the RGD linker, FT-IR spectroscopy was analyzed, and the presence of amine peaks was observed. Figure 4 shows the results of FT-IR surface analyses on bare NOA 61 and RGD-anchored NOA 61 surfaces (3 h anchoring at RT). Figure 4(a) shows the overall peaks, and Fig. 4(b) and (c) are enlarged images of (i) and (ii) in Fig. 4(a), respectively. As can be seen in Fig. 4(b), additional peaks appeared between 1500 to 1650 cm−1 on the RGD-anchored NOA 61 surface, which seem to be the peaks for amine (Venyaminov and Kalnin 1990). Figure 4(c) shows a large number of small peaks appearing alongside the curve, which are the peaks for various amino acids.
Fig. 4

FT-IR spectroscopy showing surface property of bare NOA 61 and RGD-anchored NOA 61 surfaces (3 h anchoring at RT). (a) Overall peaks. (b) Enlarged image of (i) in (a). (c) Enlarged image of (ii) in (a)

3.2.3 AFM analysis

Figure 5 shows the results of surface roughness measured using an AFM in non-contact mode. Compared to a bare NOA 61 surface that is almost flat throughout the scanned area (Fig. 5(a)), the RGD-anchored NOA 61 surface displayed a surface roughness with 20 to 200 nm in height, regardless of the anchoring temperature. However, the roughness was more regularly distributed for the RT-anchored RGD (Fig. 5(b)) compared to the 37°C-anchored RGD (Fig. 5(c)). Surface roughness accounts for the successful anchoring of the RGD linker on the NOA 61 surface.
Fig. 5

AFM images of the RGD-anchored NOA 61 layer on PET. (a) Bare NOA 61 surface. (b) RGD anchored at RT for 3 h. (c) RGD anchored at 37°C for 3 h

3.3 Cell adhesion on NOA 61-coated PET surface

Based on the surface characterization results, RGD anchoring for 3 h at RT was chosen for further study of cell adhesion experiments. Cells grew well on the bare NOA 61 surface (Fig. 6(a)); however, the RGD-anchored NOA 61 surface, performed at RT, displayed better cell adhesion compared to that on a bare NOA 61 surface, as shown in Fig. 6. Besides ECM, an RGD linker is also well-known to induce cell adhesion. Based on the cell counting, cell adhesion and growth were improved almost twice on the RGD-anchored NOA 61 surface compared to the bare NOA 61 surface. The viability of cells immobilized on the RGD-anchored NOA 61 surface was tested. Almost all cells immobilized in Fig. 6(b) were alive, represented by the green fluorescence in Fig. 6(c), and only a few cells were detected to be dead, represented by the red fluorescence in Fig. 6(d).
Fig. 6

Cell growth on (a) a bare NOA 61 and (b) RGD-anchored NOA 61 layer on PET (incubation for 3 h at RT). Fluorescence images of (c) live cells and (d) dead cells immobilized on RGD-anchored NOA 61 surfaces

3.4 Selective micropatterning of NOA on PDMS surface

NOA 61 micropatterns were successfully photopolymerized onto a PDMS surface as shown in Fig. 7. NOA 61 is a mercapto-ester—based photocurable adhesive that contains a mercapto-ester and acrylate monomers. Mercaptosilane was employed to anchor a thiol functional groups on an O2-plasma-treated PDMS surface via silane coupling reaction. Thiol groups of 3-mercaptopropyltrimethoxysilane were then reacted with the thiol groups of NOA 61 upon UV illumination, forming strong disulfide bonds for robust NOA 61 micropatterning on a PDMS surface.
Fig. 7

(a) Result of NOA 61 micropatterning on a PDMS surface and (b) its enlarged image

As can be seen in Fig. 7(a), NOA 61 micropatterns were precisely fabricated into an array format within a large area of the PDMS surface. The unpolymerized NOA 61 was removed from the PDMS surface during washing and sonication process. The robustness of the micropatterns on the PDMS surface was extremely high: The micropatterns remained undeformed on the PDMS surface even after harsh bending and stretching. They were also tolerable to a harsh sonication process in isopropyl alcohol. Without a mercaptosilane treatment, NOA 61 micropatterns were easily peeled off the PDMS surface. Although the pattern fidelity was high, the pattern size slightly increased compared to that on a photomask. This phenomenon might have resulted from the spreading of the illuminated UV light due to a relatively long distance between the photomask and PDMS surface during the photopolymerization process. However, by decreasing the distance between the mask and the surface, we could significantly decrease the pattern swelling down to 5.9%. Also, by adopting a photomask with smaller pattern sizes, relatively smaller patterns were successfully generated.

3.5 Cell adhesion and growth on NOA micropatterned PDMS surface

Figure 8(a) shows an optical microscopic (OM) image of cells immobilized on the NOA 61 micropatterned PDMS surface. As can be seen in the image, the NOA 61 micropatterns were obtained with moderate pattern fidelity, and were stable even after sonication for 12 ∼ min, representing high structural robustness. Cells were immobilized selectively on the NOA 61 micropatterns, and relatively few cells were immobilized on the surrounding (Fig. 8(a)). Figure 8(b) shows cell adhesion on a bare PDMS. Compared to those in Fig. 8(a), relatively few cells were grown on the bare PDMS surface, confirming that NOA 61 plays an important role in the selective adhesion of cells.
Fig. 8

Micropatterned NOA 61 on a PDMS surface and selective cell adhesion and growth on micropatterned NOA 61 surface. (a) OM image of cells immobilized on the NOA 61 micropatterned PDMS surface. (b) OM image of cells immobilized on a bare PDMS surface. (c) Fluorescence image of live cells immobilized on the NOA 61 micropattern. (d) Fluorescence image of dead cells

When the cells were stained with dyes, almost all the cells immobilized on NOA 61 exhibited a green fluorescence (Fig. 8(c)), indicating that most of the cells were alive. As can be seen in Fig. 8(c) and (d), the NOA 61 micropattern itself was also stained with both dyes, and fluoresced both green and red. Therefore, we can conclude that some of the dots existing on the outer boundary of the NOA 61 micropattern in Fig. 8(c) and (d) which fluoresced both green and red are NOA 61 micropatterns, and not the cells. Selective targeting of cells particularly on a PDMS surface is of great significance in microfluidic-based cell assays and detection because PDMS is the most widely adopted material for microfluidic device fabrication. Compared to conventional methods for selective cell adhesion on a PDMS adopting an ECM surface patterned using the self-assembled monolayers (SAMs) of alkanethiolates on a gold surface or alkyltrichlorosiloxanes on silicon, a NOA patterning method is relatively simple and inexpensive, and can provide guided cell alignment owing to its mechanical robustness and photocurable nature.

4 Conclusions

In summary, we have demonstrated the fabrication of robust micropattern arrays using a mercapto-ester—based photocurable adhesive on a PDMS surface mediated by mercaptosilane, and adopted a micropatterned polymer array platform for the selective patterning of cells. Due to the low viscosity and photocurable nature of NOA adhesive used, micropatterns were successfully fabricated on a chemically-inert PDMS surface, and used as a physical pathway for the targeted adhesion of cells. In addition, the effect of the cell adhesive linker, an RGD oligopeptide, was anchored for better adhesion of cells on the NOA-coated PDMS surface. Since it is difficult to precisely incorporate arrays of micropatterns directly inside a microfluidic channel, micropatterning of a cell-adhesive layer on a PDMS surface prior to forming a closed microfluidic channel is a primary concern for the construction of a cell-arrayed microfluidic system. By incorporating cell-adhesive polymer micropattern arrays on one PDMS plate and subsequently bonding it with another PDMS plate having fluidic channels, the issue of constructing a PDMS-based closed microfluidic system with a high degree of spatial selectivity for cell immobilization can be resolved, and such a system can be used for the rapid and parallel screening of drugs as well as cell-based fundamental research at the microscale level.


This work was supported by the GRRC program of Gyeonggi province (GRRC Kyungwon 2009-A01, Development of Microfluidic Devices for the Diagnosis of Disease) and the Kyungwon University Research Fund in 2009.

Copyright information

© Springer Science+Business Media, LLC 2009