Critical areas of cell adhesion on micropatterned surfaces

Critical areas of cell adhesion on micropatterned surfaces

Biomaterials 32 (2011) 3931e3938 Contents lists available at ScienceDirect Biomaterials journal homepage: www.elsevier.com/locate/biomaterials Crit...

1MB Sizes 2 Downloads 63 Views

Biomaterials 32 (2011) 3931e3938

Contents lists available at ScienceDirect

Biomaterials journal homepage: www.elsevier.com/locate/biomaterials

Critical areas of cell adhesion on micropatterned surfaces Ce Yan, Jianguo Sun, Jiandong Ding* Key Laboratory of Molecular Engineering of Polymers of Ministry of Education, Department of Macromolecular Science, Advanced Materials Laboratory, Fudan University, Shanghai 200433, China

a r t i c l e i n f o

a b s t r a c t

Article history: Received 14 January 2011 Accepted 19 January 2011 Available online 26 February 2011

The adhesive area is important to modulate cell behaviors on a substrate. This paper aims to semiquantitatively examine the existence of the characteristic areas of cell adhesion on the level of individual cells. We prepared a series of micropatterned surfaces with adhesive microislands of various sizes on an adhesion-resistant background, and cultured cells of MC3T3-E1 (osteoblast), BMSC (bone mesenchymal stem cell) or NIH3T3 (fibroblast) on those modeled surfaces. We have defined seven characteristic areas of an adhesive microisland and confirmed that they are meaningful to describe cell adhesion behaviors. Those parameters are (1) the critical adhesion area from apoptosis to survival denoted as A* or Ac1, (2) the critical area from adhesion of a single cell to adhesion of multiple cells (Ac2), (3) the basic area for one more cell to adhere (AD), (4) and (5) the characteristic areas of a microisland most probably occupied by one cell (Apeak(1)) and two cells (Apeak(2)), (6) and (7) the characteristic areas of a microisland occupied by one cell (AN(1)) or two cells (AN(2)) on average. Besides the introduction of those basic parameters, the present paper demonstrates how to determine them experimentally. We further discussed the relationship between those characteristic areas and the spreading area on a non-patterned adhesive surface. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Cell adhesion Micro-patterning Critical area Surface modification

1. Introduction Cellematrix and cellecell interactions are essential for modulating the cellular behaviors including adhesion, migration, differentiation, proliferation and apoptosis [1e12]. The corresponding understanding is thus very important for biomaterial design [13e17]. Surface patterning techniques afford a unique way to explore some fundamental cellebiomaterial interactions [18e34]. Besides topological patterns [35e42], chemical patterns with celladherent islands on a non-fouling background are very powerful to reveal some basic cellular behaviors in a well-defined geometry [43e47]. A pioneering work was published in 1997 by the groups of Ingber and Whitesides based upon observations of cells on microislands coated by fibronectin, a protein existing in extracellular matrix (ECM) on an oligo(ethylene glycol) background generated by formation of a self-assembled monolayer (SAM) via microcontact printing [48]. Many researches about cells on chemical micropatterns have sprung up since then [49e59]. One has conceptually known that the adhesive area can control cell behaviors and an appropriate adhesive microisland can achieve localization of cells. Nevertheless, some very basic questions are still open, for instance, whether or not there exists a “critical” area for adhesion of a single

* Corresponding author. Tel.: þ86 21 65643506; fax: þ86 21 65640293. E-mail address: [email protected] (J. Ding). 0142-9612/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.biomaterials.2011.01.078

cell or double cells, and if yes, how to determine the critical areas, and how about the relation of the critical areas to the spreading area of a cell on a non-patterned surface or to the project area of a cell suspending in medium, and furthermore whether or not the critical areas depend upon cell types. Herein we try to address and answer those questions. The present paper is focused upon a semi-quantitative methodological investigation of cell adhesion on a chemical micropattern. We will suggest several characteristic sizes for micropatterned surfaces. The most typical three parameters are schematically illustrated in Fig. 1. The critical adhesive area for a singe cell from apoptosis to survival is denoted as Ac1 (or A* due to its importance); the critical area of a microisland from single cell adhesion to multi-cell adhesion is defined as Ac2, which is meaningful for the control of a single cell adhesion in a potential cell chip; the characteristic area increase for adhesion of one more cell is defined as AD. According to Fig. 1, the physical picture for cells adhering onto a micropatterned surface is supposed as follows: when the adhesive area is very small (
3932

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

Fig. 1. Schematic presentation of some characteristic areas for cells adhering on microislands of different sizes. Ac1, also named A* due to its importance, the critical area from apoptosis to survival; Ac2, the critical area from single cell adhesion to multicell adhesion; AD, the characteristic area for one more cell to adhere. The arrow direction indicates the increase of area of an adhesive microisland on a cell-resistant background.

present paper aims to confirm those characteristic areas and suggest the way to determine them experimentally. These characteristic areas will, if justified and available, provide a relatively precise control of localization of cells on material surfaces. The present research must be based upon a well-defined micropatterned surface. We design an ideal model surface using arginineeglycineeaspartic acid (RGD) peptides immobilized by SeAu bond to modify the microislands and using poly(ethylene

glycol) (PEG) hydrogels as the resistant background. The RGD sequence has been identified in several adhesive proteins including fibronectin, vitronectin, laminin, collagen I, collagen IV, thrombospondin, von Willebrand factor, fibrinogen, and fibrin, and can facilitate cell adhesion in a variety of cell types by ligating to integrin, a widely expressed heterodimeric membrane receptor, which mediates cellular behaviors or functions as a linker between ECM and cytoskeleton [1,60e62]. Thus RGD peptides have been widely employed to tune cellematerial interactions by many researchers [63e69]. Compared to replacement of fibronectin proteins with RGD peptides, the substituent of a SAM of oligo (ethylene glycol) by a PEG hydrogel is more important for the present study. Patterning cells onto isolated microdomains demands an excellent non-fouling background to avoid adsorption of proteins in the culture medium and thus of cells. While the SAM of oligo(ethylene glycol) might detach from the gold surface, a PEG hydrogel can repel cell adhesion better and much longer [70]. To this end, we will fabricate patterns of gold microislands on PEG hydrogels via a transfer photolithography technique with the basic

Fig. 2. Schematic presentation of the process of fabrication of an RGD micropattern on a PEG background. Three basic stages are included: (a) fabrication of gold micropatterns on glass via lift-off photolithography; (b) transfer of gold patterns from glass (hard) to a PEG hydrogel (soft); (c) formation of SAMs of RGD peptides on gold micropattern via postmodification by c(-RGDfK-)-thiol ligands.

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

stages shown in Fig. 2. A series of RGD microislands of different sizes on PEG hydrogels will be generated, and cells will be cultured on these microislands in order to examine the characteristic areas of cell adhesion on micropatterned surfaces. 2. Materials and methods

3933

Co.) were casted, and crosslinked under 365 nm ultraviolet light; last, the PEG hydrogel was peeled off to get hydrogel surface with gold microislands. In the last stage, PEG hydrogel were swelled and grafted with RGD peptides via a self-assembly of c(-RGDfK-)-thiol ligands (R: arginine, G: glycine, D: aspartic acid, f: D-phenylalanine, and K: lysine) (Peptides International, USA) of 25 mM on gold micropatterns at 4  C for 8 h. A micropattern of RGD microislands on the PEG hydrogel surface was finally obtained (Fig. 2c).

2.1. The preparation of RGD micropatterns on PEG hydrogels 2.2. Cell culture The fabrication strategy of the model micropatterned surface consists of three stages (Fig. 2). The first stage is to fabricate gold micropatterns on glass slides via liftoff photolithography. The glass slides were cleaned by “piranha” solution [H2SO4/ 30% H2O2, 3:1 (v/v)] (Caution: be careful when carrying out this violent reaction.), sonicated by Milli-Q water, and baked. The cleaned glass slides were spin-coated with a layer of positive photoresist RZJ-304 (Suzhou Ruihong Electronic Chemicals Co. Ltd., China), baked, exposed to ultraviolet light through a pre-designed chrome mask, treated by corresponding developer RZX-3038 (Suzhou Ruihong Electronic Chemicals Co. Ltd., China), and baked once more. Then, a layer of gold was sputtered onto glass slides with micropatterns of photoresist following oxygen plasma treatment. After the photoresist was lifted-off by acetone, we got gold micropatterns on glass slides (Fig. 2a). A transfer technique [71] was then used to generate gold microislands on PEG hydrogels (Fig. 2b). First, allyl mercaptan (Fluka) was grafted onto gold micropatterns via SeAu bond in vacuum; then macromonomers of poly(ethylene glycol) diacrylate (PEG-DA, Mn 700, Sigma) mixed with 0.05 wt% photoinitiator 2-hydroxy1-[4-(hydroxyethoxy) phenyl]-2-methyl-1-propanone (D2959, Aldrich Chemical

MC3T3-E1 cells (mouse osteoblastic cell line) and NIH-3T3 cells (mouse embryonic fibroblast cell line) were purchased from Shanghai Cell Bank in China. MC3T3-E1 cells were cultured in minimum essential medium alpha (MEMa, Gibco) plus 10% fetal bovine serum (FBS, Biochrom) and NIH-3T3 cells in high-glucose Dulbecco’s modified Eagle medium (DMEM) plus 10% FBS (Gibco). Bone marrow stromal cells (BMSCs) were isolated from newborn neonatal Sprague Dawley (SD) rats, cultured in low-glucose DMEM (Gibco) supplemented with 10% FBS (Gibco), and the second passage was used for cell culture. All the culture media were supplemented with 100 U/ml penicillin, 100 mg/ml streptomycin and 2 mM Lglutamine (Gibco). Prior to use, the PEG hydrogel with RGD micropatterns were sterilized by 75% alcohol, exhaustively rinsed by phosphate buffered saline (PBS) solution and pressed by polytetrafluorethylene tubes with inner diameter about 15 mm. All of the three cell lines were plated at three densities including 2, 4, 6  104 cells/well in 12-well plates in the corresponding culture media at 37  C with 5% CO2 for 8 h. Those non-adherent cells were removed by changing culture medium after 1 h culture.

Fig. 3. (a) Optical micrographs of as-fabricated micropatterns of RGD-grafted gold microislands on a PEG hydrogel. We generated microislands of 25 diameters including 4, 6, 8, 10, 12, 14, 16 mm (line 4), 18, 20, 22, 24, 26, 28, 30 mm (line 2), 32, 35, 40, 45, 50, 55 mm (line 3), and 60, 70, 80, 90, 100 mm (line 1). (b) A typical phase contrast micrograph of MC3T3-E1 cells on the micropatterned surface demonstrating the excellent localization of cells on the adhesive microislands. (c) A fluorescence image of MC3T3-E1 cells by viability staining illustrating good cell viability of the adherent cells (d, e) Fluorescent micrographs with cellular nucleus stained as blue (d) and also F-actin stained as red (e) showing the possibility of determination of the number of adherent cells on each microisland. The dashed circles here indicate the contour of the underlying RGD-grafted microislands. All the bars are 50 mm. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3934

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

2.3. Observations of cell adhesion on micropatterned surfaces In order to guarantee cell adhesion while avoiding cell proliferation, cells were observed at 8 h post-seeding. Then, cells were fixed by 4% paraformaldehyde, and washed in PBS. The nucleuses were stained with 40 , 6-diamidino-2-phenylindole (DAPI, Sigma) for cell number counting. For detecting the organization of F-actin, cells on PEG hydrogels were fixed at 8 h, permeabilized with 0.1% Triton-X 100 in PBS, incubated with phalloidin-tetramethylrhodamine B isothiocyanate (phalloidin-TRITC, Sigma), rinsed with PBS, stained with DAPI, and rinsed by PBS and distilled water again. An inverted microscope (Axiovert 200, Zeiss) mounted with a Canon digital camera was used to take images. 2.4. Cell viability assay In order to check the viability of cells on micropatterned RGD islands, cells cultured for 8 h were hatched with LIVE/DEADÒ viability/cytotoxicity assay kit to simultaneously determine live and dead cells. Live cells were labeled as green by calcein AM and dead cells as red by ethidium homodimer-1. 2.5. Measurements of project areas of spreading and suspended cells In order to get the spreading area of cells freely adhering without pattern constraint, the above cell lines were also seeded onto a layer of gold grafted with c (-RGDfK-)-thiol ligands on glass slides at the density of 4  104 cells/well in six-well plates, and cultured for 8 h. A thin transition monolayer of chromium was evaporated prior to gold deposition for substrate adherence [72]. Cells cultured on RGDgrafted gold layers were stained with phalloidin-TRITC (F-actin) and DAPI (nucleus) to evaluate the project area of spreading cells. Suspended cells obtained by digestion with 0.25% trypsin-EDTA (Gibco) were plated onto tissue culture plates (TCPs), and immediately observed under a microscope. Snapshots of suspended cells just falling onto substrate surfaces were taken at bright field.

Fig. 4. Fraction of cellular occupation per microisland (fN>0, triangles) and fraction of occupation of multiple cells (fN>1, inverted triangles) as a function of adhesive area of microislands A. The critical areas A* and Ac2 are defined as the crosspoints between the corresponding two marked asymptote lines.

2.6. Data analysis The project areas of cells freely adhering on RGD-grafted gold layers (Aspread) were gained by dealing with micrographs of cells after F-actin staining using free software Image-J (freely available at http://www.nih.gov), and those of suspended cells (Asuspend) were measured via the micrographs taken at bright field. We prepared 3 samples for averaging. At least 150 spreading cells and at least 100 suspended cells were randomly collected to determine the average project area for each sample. The data were analyzed by student’s t-test. The statistical significance was accepted when p < 0.05.

3. Results and discussion 3.1. Whether those “characteristic areas” exist for cells on a micropatterned surface? To evaluate the existence of the “defined” characteristic areas such as Ac1 (also denoted as A*, the critical area from apoptosis to survival), and Ac2 (the critical area from single cell adhesion to multi-cell adhesion), we first tried to observe and analyze MC3T3E1 cells with a loading density of 2  104 cells/well on micropatterned surfaces (Fig. 3). Comparison of Fig. 3b with Fig. 3a indicated that our model surface enabled an excellent localization of cells on given adhesive microislands. We further examined the viability of those adherent cells. Under fluorescent observations, live cells show green fluorescence and dead cells show red fluorescence. Most of MC3T3-E1 cells adhering on RGD microislands including those small microislands exhibited good viability (Fig. 3c). Cell nucleuses were stained by DAPI for cell number counting (Fig. 3d). One intact nuclear represents an adherent cell. With the increase of the adhesive area as pointed out by the dashed circles, the cell number increased and the cell morphology turned from confined to spreading (Fig. 3e). This is basically consistent with the picture shown in Fig. 1. A microisland with either one cell or multiple cells is regarded as “occupied”. The occupation fraction fN>0 (subscript N indicating cell number) is defined as the percentage of occupied islands out of all statistical units. fN>0 (scattered solid triangles) as a function of

Fig. 5. (a) Average cell number versus area of adhesive microisland for MC3T3-E1 cells. The area with an average cell number equal to 1 is labeled as AN(1), and AN(2) represents the area when the average cell number is 2. The red solid line is the linear fit result according to the last 10 experimental points. The reciprocal of the slope is defined as AD (the area for one more cell to adhere). (b) Local cell density on microislands d as indicated by the number of cells per 1000 mm2 of adhesive microislands as a function of microisland area A in a logarithmic scale. For convenience of comparison, Ac1 (A*) and Ac2 are also marked. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

adhesive area is shown in Fig. 4. It is necessary to note that the linear swelling ratio of gold microislands on PEG hydrogels in an aqueous environment measured as 14% has been taken into consideration in calculation of the actual areas of those microislands. When the island area was sufficiently small, the probability of cells to adhere on the microisland was zero. As the adhesive area increased, fN>0 gradually increased from zero to unity, which demonstrated a transition from “non-adhesive” to “adhesive” microisland, or implied a switch from cell “apoptosis” to “survival”. This size dependence of cell adhesion is in accordance with reference [48]. The crosspoint of two tangential lines was employed to determine the critical point in Fig. 4. In this way, we obtained the critical area from apoptosis to survival Ac1 (also named A*), similar to the determination of critical micelle concentration (CMC) from the OD-concentration curve in studies of the micellization transition of amphiphiles in a selective solvent [73,74]. According to Fig. 4, A* ¼ 64 mm2. For the convenience of comparison of characteristic areas of cells on patterned and non-patterned surfaces, we also measured the corresponding project area of spreading cells on a nonpatterned surface (Aspread) and suspended cells (Asuspend) using software Image-J. The measurement of Aspread was based on the fluorescence images of cells with F-actin stained as red (Fig. S1).

Fig. 6. (a) Population fraction of MC3T3-E1 cells with Ncell ¼ 0, 1, 2, 3, 4e5, 6e8, 9e12, and >12 for each microisland at a seed density of 20,000 cells per well. The peak values for Ncell ¼ 1 and 2 are represented as Apeak(1) and Apeak(2), respectively. (b) The distribution of cell number on some special adhesive areas approximate to Ac1, Ac2, AN(1), AN(2), Apeak(1) and Apeak(2).

3935

The project area of MC3T3-E1 cells on an RGD-grafted gold-layer 8 h after seeding was determined as Aspread ¼ 2694  81 mm2. Most cells have spread to normal sizes with polygonal morphologies and mature skeleton. Asuspend was analyzed due to bright field images with clear morphology boundary (also Fig. S1), and had a value of 325  1 mm2. Ac1 is much smaller than the average project area of either freely spreading or suspended cells. This is not hard to understand that the cells on patterned surface are constrained, especially on those small microislands. It also reveals that a confined morphology tolerates a smaller adhesive area than its actual project area (the second image in Fig. 3e). Cell’s adhering on small microislands in Fig. 3b and c also supports this viewpoint comparing with sizes of adhesive microislands in Fig. 3a. Such constrained single cells probably posses a horizontally compressed and vertically heightened shape to keep its volume (Fig. 1). Similarly, we defined an island with at least two cells as “multioccupied”. The fractional multi-occupation fN>1 denotes the percentage of multi-occupied islands, and the corresponding data under varying areas of microislands are shown in Fig. 4. The curve of fN>1 versus adhesive area (A) also exhibited a similar transition resembling CMC. So, the critical area from a single cell to multiple cells Ac2 could be determined as Ac2 ¼ 335 mm2. In the same way, the fraction of N  n (called fN>n) could be obtained (data not shown), and for example Ac3 ¼ 662 mm2.

Fig. 7. (a) Fraction of cell occupation on microislands (fN>0) as a function of adhesive area of microisland for indicated cell types. The lines shown in (a) are used to determine A* values from the crosspoints. (b) The experimentally determined critical areas A*, Ac2 and AD for MC3T3-E1, BMSC and NIH3T3 cells. The seed density was 20, 000 cells per well.

3936

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

We also calculated the average cell number on microislands (Fig. 5a). When the adhesive area was sufficiently large, the cell number increased almost linearly, which illustrated the existence of a given area for one more cell to adhere. We defined this characteristic area as AD, which could be easily obtained from the reciprocal value of the slope of the linear fit. AD resulted in 1058  22 mm2. This value was larger than Asuspend, but smaller than Aspread, which indicated a crowded state on adhesive microislands on a resistant background. The areas with the average cell number equal to 1 and 2 are labeled as AN(1) and AN(2), respectively. According to Fig. 5a, AN(1) ¼ 643 mm2, and AN(2) ¼ 1285 mm2. It is not surprising that AN(1) and AN(2) are significantly larger than Ac1 and Ac2, respectively. We further calculated the local density of MC3T3-E1 cells on each microisland (Fig. 5b). The most crowded state happened when the area of adhesive microislands was between Ac1 and Ac2, and relatively close to Ac2. Since the cell number on each island area was polydispersed, we checked the distribution of cell number on each adhesive area. The populations of special cell numbers like 0, 1, 2 on different adhesive microislands are plotted in Fig. 6a. We defined the areas with a maximum population for a given cell number 1 or 2 as Apeak(1) and Apeak(2), respectively. According to Fig. 6a, Apeak(1) ¼ 623 mm2 and Apeak(2) ¼ 874 mm2. The distributions of cell number on microislands with areas of 65, 331, 588, 919, 690, and 1250 mm2 approximate to A* (64 mm2), Ac2 (335 mm2), Apeak(1) (623 mm2), Apeak(2) (874 mm2), AN(1) (643 mm2) and AN(2) (1285 mm2) are plotted in Fig. 6b. The population of Ncell ¼ 0 is equal to 1  fN>0. The cell number on microislands with area around A* showed an on/off (1 or 0) distribution, and the corresponding fN>0 was about 10% (0.091  0.036). The fraction of occupation increased significantly with microisland area. 3.2. The ubiquity of the concept of the characteristic areas In order to confirm the justification of the concept of the above characteristic areas, we further examined two other cell types, a fibroblast NIH3T3 and a stem cell BMSC besides MC3T3-E1. The

Table 1 Average ratios of the characteristic areas over spreading area. Ratio over Aspread

Mean

Range

A* Ac2 Apeak(1) Apeak(2) AN(1) AN(2) AD Asuspend

0.03 0.14 0.18 0.35 0.27 0.50 0.30 0.11

0.01e0.08 0.06e0.21 0.09e0.27 0.19e0.51 0.14e0.48 0.32e0.72 0.21e0.39 0.10e0.12

MC3T3-E1 is a well-characterized murine osteoblast cell line and an established model for osteoblasts [75]. BMSCs are stem cells which can differentiate in multiple directions including osteoblast, chondrocyte, and adipocyte lineages and so on [49]. In conventional biological knowledge, NIH3T3 fibroblasts have no differentiation potential; however, the emergence of iPS cells sheds new light of availability of differentiation of cells including fibroblasts [76]. These three cell types are different in morphologies and functions, and all of them are very popular. After cell culture and staining, we carried out the same statistical process for both BMSCs and NIH3T3 cells. The Asuspend and Aspread values were measured as 297  11 and 3021  151 mm2 for BMSCs, and 265  12 and 2200  79 mm2 for NIH3T3 cells. The above-defined characteristic areas also can be applied to BMSCs and NIH3T3 cells. The acquisitions of A* for three cell lines were plotted together to show the ubiquity of the concept and determination algorithm for A* (Fig. 7a). A summary of the three characteristic areas for the three cell types is shown in Fig. 7b. While the concept of the critical areas for cell adhesion on micropatterned surfaces is universal, each cell type has its own characteristic values. For Ac1 (or A*), MC3T3-E1 < BMSC < NIH3T3. This implies that the MC3T3-E1 cells may demand a lowest number or highest strength of single focal adhesion to allow adherence on very small microislands. For the Ac2 value, the relationship was MC3T3-E1 < NIH3T3 < BMSC, and thus it is relatively easy for MC3T3-E1 cells to accept a second cell on small microislands. As AD is concerned, MC3T3-E1 > BMSC z NIH3T3. For the crowded state

Fig. 8. A summarized presentation of all characteristic areas for three cell types: MC3T3-E1 cells (red), BMSCs (green), NIH3T3 cells (blue) at density of 20, 000 cells/well. A*, also named Ac1 is the critical area from apoptosis to survival; Ac2, the critical area from single- to multi-cell adhesion; AD, the area for one more cell to adhere; Apeak(1), the area with respect to the maximum population among all microislands occupied exactly by a single cell; Apeak(2), the area with respect to the maximum population among all microislands occupied exactly by two cells; AN(1), the area of the microisland occupied by one cell on average; AN(2), the area of the microisland occupied by two cells on average; Aspread, the averaged project area of freely spreading cells on a non-patterned adhesive surface; Asuspend, the averaged project area of suspended cells. They are sorted into three groups for a clear vision: Ac1, Apeak(1) and AN(1); Ac2, Apeak(2) and AN(2); Asuspend, AD and Asuspend. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

of a multi-cell cluster, MC3T3-E1 cells need a largest area to add one cell probably due to its preference to a large cell volume. It should be indicated that due to the compressibility of cells, density of cell seeding may affect the average cell number on an island of a given adhesive area, and thus influence the concentration of growth factors and the lowest survival adhesive area Ac1 (or A*). The corresponding characteristic areas for three cell lines at three seed densities of 2, 4 and 6  104 cells/well are listed in Table S1 in the supplementary information. 3.3. The relationship between those characteristic areas Although the characteristic areas are dependent upon cell types and also on some seeding conditions, we found that some relations between those areas always maintained. Fig. 8 shows all of the above-mentioned characteristic areas for three cell types. It clearly reveals that A* (or Ac1) << Asuspend, Ac2 < Apeak(1) < AN(1), AD, Apeak(2) < AN(2) << Aspread for all of the cell types cultured in this paper. Although the characteristic areas related to micropatterns decreased with cell density in the range examined in this paper, these relationships are justified for all densities (Table S1). The ratio of some characteristic areas including A*/Asuspend, A*/ Aspread, A*/Ac2, Ac2/Aspread and AD/Aspread are summarized in Table S2. In order to clearly elucidate these ratios, the characteristic ratios were averaged over cell type (MC3T3-E1, BMSC, NIH3T3) or seed density (2, 4, 6  104 cells/well), and over all cases (9 in total, Table S3). The average ratios involving A* and Ac2 are more sensitive to cell type than seed density, while AD is more related to seed densities. We account for the above trends as follows: stronger adhesion capability of a cell type leads to easier adhesion on substrate and thus smaller A* and Ac2, and a high seed density causes crowding and thus a decrease of AD. In order to roughly indicate the universal relations between those seven characteristic areas of cells on micropatterns defined in this paper and also the other two normal basic parameters of cells Asuspend and Aspread, we summarized the ratios of those characteristic areas over Aspread in all cases (Table 1). If we set Aspread as 100 and the ratios are simplified in reduced units of 5 in an easymemory way, the rough values for the 9 areas could be simply summarized as: A* or Ac1, 5; Asuspend, 10; Ac2, 15; Apeak(1), 20; AN(1), 25; AD, 30; Apeak(2), 35; AN(2), 50; Aspread: 100. According to our experimental observations in this paper, A* is much smaller than Asuspend. We presumed that this critical adhesive area might be determined by a minimum necessary adhesion force rather than the adhesive area. As a result, these characteristic areas may be affected by cellesubstrate adhesive ability like ligand density, chemical composition (e.g. other ligands like collagen) of adhesive regions, substrate modulus and so on. In most cases, Ac2 values were found to be much larger than two folds of Ac1, and much smaller than two folds of Asuspend. Thus, Ac2 might be mainly related to total adhesive force and the spatial hindrance between two adjacent cells. 4. Conclusions Seven characteristic areas for cell adhesion on micropatterned surfaces with a resistant background have been introduced, including the critical adhesion area from apoptosis to survival (A* or Ac1), the critical adhesion area from occupation of a single cell to multiple cells (Ac2), and the basic adhesion area for one more cell to adhere on a microisland (AD). A series of micropatterns of high adhesion contrast containing RGD microislands of varied but welldefined sizes on PEG hydrogels were fabricated, and three types of cells, MC3T3-E1, BMSC and NIH3T3 were cultured on those micropatterned surfaces. The present paper has justified the concept of

3937

critical areas to describe semi-quantitatively the basic size-dependent behaviors of cell adhesion on a micropatterned substrate, and put forward the practical approaches of their determination. The concept has been confirmed to be universal in all of the three cell types, and meanwhile those characteristic areas are dependent upon cell types as well as influenced by some experimental conditions such as seed density. Some general relationships between these characteristic areas have further been revealed as A* (or Ac1) << Asuspend, Ac2 < Apeak(1) < AN(1), AD, Apeak(2) < AN(2) << Aspread. Hence, this work provides the fundamentals for size design of micropatterns to control localization of a given number of cells, and seems also stimulating for some other pertinent fields like cell chips and bio-modelings. Acknowledgments The authors are grateful for the financial supports from Chinese Ministry of Science and Technology (973 programs No. 2009CB930000 and No. 2011CB606203), and NSF of China (grant No. 21034002). Appendix. Supplementary material Supplementary data related to this article can be found online at doi:10.1016/j.biomaterials.2011.01.078. References [1] Geiger B, Spatz JP, Bershadsky AD. Environmental sensing through focal adhesions. Nat Rev Mol Cell Biol 2009;10:21e33. [2] Discher DE, Janmey P, Wang YL. Tissue cells feel and respond to the stiffness of their substrate. Science 2005;310:1139e43. [3] Grigoriou V, Shapiro IM, Cavalcanti-Adam EA, Composto RJ, Ducheyne P, Adams CS. Apoptosis and survival of osteoblast-like cells are regulated by surface attachment. J Biol Chem 2005;280:1733e9. [4] Nagaoka M, Koshimizu U, Yuasa S, Hattori F, Chen H, Tanaka T, et al. E-cadherin-coated plates maintain pluripotent ES cells without colony formation. PLoS One 2006;1:1e7. [5] Polte TR, Shen MY, Karavitis J, Montoya M, Pendse J, Xia S, et al. Nanostructured magnetizable materials that switch cells between life and death. Biomaterials 2007;28:2783e90. [6] Su J, Jiang X, Welsch R, Whitesides GM, So PTC. Geometric confinement influences cellular mechanical properties I e adhesion area dependence. Mol Cell Biomech 2007;4:87e104. [7] Khetan S, Burdick JA. Patterning network structure to spatially control cellular remodeling and stem cell fate within 3-dimensional hydrogels. Biomaterials 2010;31:8228e34. [8] Tang ZL, Akiyama Y, Yamato M, Okano T. Comb-type grafted poly(N-isopropylacrylamide) gel modified surfaces for rapid detachment of cell sheet. Biomaterials 2010;31:7435e43. [9] Pek YS, Wan ACA, Ying JY. The effect of matrix stiffness on mesenchymal stem cell differentiation in a 3D thixotropic gel. Biomaterials 2010;31:385e91. [10] Gentile F, Tirinato L, Battista E, Causa F, Liberale C, di Fabrizio EM, et al. Cells preferentially grow on rough substrates. Biomaterials 2010;31:7205e12. [11] Mercey E, Obeïd P, Glaise D, Calvo-Munoz M-L, Guguen-Guillouzo C, Fouqué B. The application of 3D micropatterning of agarose substrate for cell culture and in situ comet assays. Biomaterials 2010;31:3156e65. [12] Nguyen LH, Kudva AK, Guckert NL, Linse KD, Roy K. Unique biomaterial compositions direct bone marrow stem cells into specific chondrocytic phenotypes corresponding to the various zones of articular cartilage. Biomaterials 2011;32:1327e38. [13] Curran JM, Chen R, Hunt JA. The guidance of human mesenchymal stem cell differentiation in vitro by controlled modifications to the cell substrate. Biomaterials 2006;27:4783e93. [14] Williams DF. On the mechanisms of biocompatibility. Biomaterials 2008;29: 2941e53. [15] Williams DF. On the nature of biomaterials. Biomaterials 2009;30:5897e909. [16] Yue XS, Murakami Y, Tamai T, Nagaoka M, Cho CS, Ito Y, et al. A fusion protein N-cadherin-Fc as an artificial extracellular matrix surface for maintenance of stem cell features. Biomaterials 2010;31:5287e96. [17] Scharnagl N, Lee S, Hiebl B, Sisson A, Lendlein A. Design principles for polymers as substratum for adherent cells. J Mater Chem 2010;20:8789e802. [18] den Braber ET, de Ruijter JE, Ginsel LA, von Recum AF, Jansen JA. Orientation of ECM protein deposition, fibroblast cytoskeleton, and attachment complex components on silicone microgrooved surfaces. J Biomed Mater Res 1998;40: 291e300.

3938

C. Yan et al. / Biomaterials 32 (2011) 3931e3938

[19] Balaban NQ, Schwarz US, Riveline D, Goichberg P, Tzur G, Sabanay I, et al. Force and focal adhesion assembly: a close relationship studied using elastic micropatterned substrates. Nat Cell Biol 2001;3:466e72. [20] Falconnet D, Csucs G, Grandin HM, Textor M. Surface engineering approaches to micropattern surfaces for cell-based assays. Biomaterials 2006;27:3044e63. [21] Charest JL, García AJ, King WP. Myoblast alignment and differentiation on cell culture substrates with microscale topography and model chemistries. Biomaterials 2007;28:2202e10. [22] Yap FL, Zhang Y. Assembly of polystyrene microspheres and its application in cell micropatterning. Biomaterials 2007;28:2328e38. [23] Biggs MJP, Richards RG, Gadegaard N, Wilkinson CDW, Oreffo ROC, Dalby MJ. The use of nanoscale topography to modulate the dynamics of adhesion formation in primary osteoblasts and ERK/MAPK signalling in STRO-1þ enriched skeletal stem cells. Biomaterials 2009;30:5094e103. [24] Yao L, Wang SG, Cui WJ, Sherlock R, O’Connell C, Damodaran G, et al. Effect of functionalized micropatterned PLGA on guided neurite growth. Acta Biomater 2009;5:580e8. [25] Chien H-W, Chang TY, Tsai W- B. Spatial control of cellular adhesion using photo-crosslinked micropatterned polyelectrolyte multilayer films. Biomaterials 2009;30:2209e18. [26] Cortese B, Piliego C, Viola I, D’Amone S, Cingolani R, Gigli G. Engineering transfer of micro- and nanometer-scale features by surface energy modification. Langmuir 2009;25:7025e31. [27] Huang J, Gräter SV, Corbellini F, Rinck S, Bock E, Kemkemer R, et al. Impact of order and disorder in RGD nanopatterns on cell adhesion. Nano Lett 2009;9:1111e6. [28] Prodanov L, te Riet J, Lamers E, Domanski M, Luttge R, van Loon JJWA, et al. The interaction between nanoscale surface features and mechanical loading and its effect on osteoblast-like cells behavior. Biomaterials 2010;31:7758e65. [29] Curran JM, Stokes R, Irvine E, Graham D, Amro NA, Sanedrin RG, et al. Introducing dip pen nanolithography as a tool for controlling stem cell behaviour: unlocking the potential of the next generation of smart materials in regenerative medicine. Lab Chip 2010;10:1662e70. [30] Chen S, Jones JA, Xu Y, Low H-Y, Anderson JM, Leong KW. Characterization of topographical effects on macrophage behavior in a foreign body response model. Biomaterials 2010;31:3479e91. [31] Decuzzi P, Ferrari M. Modulating cellular adhesion through nanotopography. Biomaterials 2010;31:173e9. [32] Rape AD, Guo WH, Wang YL. The regulation of traction force in relation to cell shape and focal adhesions. Biomaterials 2011;32:2043e51. [33] Kaur G, Valarmathi MT, Potts JD, Jabbari E, Sabo-Attwood T, Wang Q. Regulation of osteogenic differentiation of rat bone marrow stromal cells on 2D nanorod substrates. Biomaterials 2010;31:1732e41. [34] Vianay B, Kafer J, Planus E, Block M, Graner F, Guillou H. Single cells spreading on a protein lattice adopt an energy minimizing shape. Phys Rev Lett 2010;105:128101. [35] Yim EKF, Darling EM, Kulangara K, Guilak F, Leong KW. Nanotopography-induced changes in focal adhesions, cytoskeletal organization, and mechanical properties of human mesenchymal stem cells. Biomaterials 2010;31:1299e306. [36] Lamers E, Walboomers XF, Domanski M, te Riet J, van Delft FCMJM, Luttge R, et al. The influence of nanoscale grooved substrates on osteoblast behavior and extracellular matrix deposition. Biomaterials 2010;31:3307e16. [37] Guvendiren M, Burdick JA. The control of stem cell morphology and differentiation by hydrogel surface wrinkles. Biomaterials 2010;31:6511e8. [38] Wan YQ, Wang Y, Liu ZM, Qu X, Han BX, Bei JZ, et al. Adhesion and proliferation of OCT-1 osteoblast-like cells on micro- and nano-scale topography structured poly(L-lactide). Biomaterials 2005;26:4453e9. [39] Zhu J, Li J, Wang B, Zhang WJ, Zhou G, Cao Y, et al. The regulation of phenotype of cultured tenocytes by microgrooved surface structure. Biomaterials 2010;31:6952e8. [40] Bettinger CJ, Orrick B, Misra A, Langer R, Borenstein JT. Microfabrication of poly (glycerol-sebacate) for contact guidance applications. Biomaterials 2006;27:2558e65. [41] Kim DH, Han K, Gupta K, Kwon KW, Suh KY, Levchenko A. Mechanosensitivity of fibroblast cell shape and movement to anisotropic substratum topography gradients. Biomaterials 2009;30:5433e44. [42] Moeller H-C, Mian MK, Shrivastava S, Chung BG, Khademhosseini A. A microwell array system for stem cell culture. Biomaterials 2008;29:752e63. [43] Kane RS, Takayama S, Ostuni E, Ingber DE, Whitesides GM. Patterning proteins and cells using soft lithography. Biomaterials 1999;20:2363e76. [44] Patrito N, McCague C, Norton PR, Petersen NO. Spatially controlled cell adhesion via micropatterned surface modification of poly(dimethylsiloxane). Langmuir 2007;23:715e9. [45] Jang K, Sato K, Mawatari K, Konno T, Ishihara K, Kitamori T. Surface modification by 2-methacryloyloxyethyl phosphorylcholine coupled to a photolabile linker for cell micropatterning. Biomaterials 2009;30:1413e20. [46] Sun JG, Tang J, Ding JD. Cell orientation on a stripe-micropatterned surface. Chin Sci Bull 2009;54:3154e9. [47] Ahmed WW, Wolfram T, Goldyn AM, Bruellhoff K, Rioja BA, Möller M, et al. Myoblast morphology and organization on biochemically micro-patterned hydrogel coatings under cyclic mechanical strain. Biomaterials 2010;31: 250e8.

[48] Chen CS, Mrksich M, Huang S, Whitesides GM, Ingber DE. Geometric control of cell life and death. Science 1997;276:1425e8. [49] McBeath R, Pirone DM, Nelson CM, Bhadriraju K, Chen CS. Cell shape, cytoskeletal tension, and RhoA regulate stem cell lineage commitment. Dev Cell 2004;6:483e95. [50] Théry M, Racine V, Pépin A, Piel M, Chen Y, Sibarita JB, et al. The extracellular matrix guides the orientation of the cell division axis. Nat Cell Biol 2005;7:947e53. [51] Théry M, Pépin A, Dressaire E, Chen Y, Bornens M. Cell distribution of stress fibres in response to the geometry of the adhesive environment. Cell Motil Cytoskeleton 2006;63:341e55. [52] Dusseiller MR, Schlaepfer D, Koch M, Kroschewski R, Textor M. An inverted microcontact printing method on topographically structured polystyrene chips for arrayed micro-3-D culturing of single cells. Biomaterials 2005;26:5917e25. [53] Nalayanda DD, Kalukanimuttam M, Schmidtke DW. Micropatterned surfaces for controlling cell adhesion and rolling under flow. Biomed Microdevices 2007;9:207e14. [54] Ingber DE. Tensegrity-based mechanosensing from macro to micro. Prog Biophys Mol Biol 2008;97:163e79. [55] Shen CJ, Fu JP, Chen CS. Patterning cell and tissue function. Cell Mol Bioeng 2008;1:15e23. [56] Kalinina S, Gliemann H, Lopez-Garcia M, Petershans A, Auernheimer J, Schimmel T, et al. Isothiocyanate-functionalized RGD peptides for tailoring cell-adhesive surface patterns. Biomaterials 2008;29:3004e13. [57] Kilian KA, Bugarija B, Lahn BT, Mrksich M. Geometric cues for directing the differentiation of mesenchymal stem cells. Proc Natl Acad Sci U S A 2010;107: 4872e7. [58] Elloumi-Hannachi I, Maeda M, Yamato M, Okano T. Portable microcontact printing device for cell culture. Biomaterials 2010;31:8974e9. [59] Tuleuova N, Lee JY, Lee J, Ramanculov E, Zern MA, Revzin A. Using growth factor arrays and micropatterned co-cultures to induce hepatic differentiation of embryonic stem cells. Biomaterials 2010;31:9221e31. [60] García AJ, Boettiger D. Integrinefibronectin interactions at the cell-material interface: initial integrin binding and signaling. Biomaterials 1999;20: 2427e33. [61] García AJ. Get a grip: integrins in cellebiomaterial interactions. Biomaterials 2005;26:7525e9. [62] Siebers MC, ter Brugge PJ, Walboomers XF, Jansen JA. Integrins as linker proteins between osteoblasts and bone replacing materials. A critical review. Biomaterials 2005;26:137e46. [63] Kantlehner M, Schaffner P, Finsinger D, Meyer J, Jonczyk A, Diefenbach B, et al. Surface coating with cyclic RGD peptides stimulates osteoblast adhesion and proliferation as well as bone formation. ChemBioChem 2000;1:107e14. [64] Lee MH, Adams CS, Boettiger D, DeGrado WF, Shapiro IM, Composto RJ, et al. Adhesion of MC3T3-E1 cells to RGD peptides of different flanking residues: detachment strength and correlation with long-term cellular function. J Biomed Mater Res Part A 2007;81A:150e60. [65] Tugulu S, Silacci P, Stergiopulos N, Klok HA. RGD e functionalized polymer brushes as substrates for the integrin specific adhesion of human umbilical vein endothelial cells. Biomaterials 2007;28:2536e46. [66] Miller JS, Shen CJ, Legant WR, Baranski JD, Blakely BL, Chen CS. Bioactive hydrogels made from step-growth derived PEG-peptide macromers. Biomaterials 2010;31:3736e43. [67] Huang JH, Ding JD. Nanostructured interfaces with RGD arrays to control cellematrix interaction. Soft Matter 2010;6:3395e401. [68] Lai Y, Xie C, Zhang Z, Lu W, Ding JD. Design and synthesis of a potent peptide containing both specific and non-specific cell-adhesion motifs. Biomaterials 2010;31:4809e17. [69] Zhang Z, Lai YX, Yu L, Ding JD. Effects of immobilizing sites of RGD peptides in amphiphilic block copolymers on efficacy of cell adhesion. Biomaterials 2010;31:7873e82. [70] Tang J, Peng R, Ding JD. The regulation of stem cell differentiation by cellecell contact on micropatterned material surfaces. Biomaterials 2010;31:2470e6. [71] Sun JG, Graeter SV, Yu L, Duan SF, Spatz JP, Ding JD. Technique of surface modification of a cell-adhesion-resistant hydrogel by a cell-adhesion-available inorganic microarray. Biomacromolecules 2008;9:2569e72. [72] Moody NR, Adams DP, Volinsky AA, Kriese MD, Gerberich WW. Annealing effects on interfacial fracture of goldechromium films in hybrid microcircuits. In: Carter CB, Hall EL, Nutt SR, Briant CL, editors. Interfacial engineering for optimized properties, vol. II. Warrendale: Materials Research Society; 2000. p. 195e206. [73] Yu L, Zhang Z, Zhang H, Ding JD. Mixing a sol and a precipitate of block copolymers with different block ratios leads to an injectable hydrogel. Biomacromolecules 2009;10:1547e53. [74] Chang GT, Yu L, Yang ZG, Ding JD. A delicate ionizable-group effect on selfassembly and thermogelling of amphiphilic block copolymers in water. Polymer 2009;50:6111e20. [75] Chatterjee K, Lin-Gibson S, Wallace WE, Parekh SH, Lee YJ, Cicerone MT, et al. The effect of 3D hydrogel scaffold modulus on osteoblast differentiation and mineralization revealed by combinatorial screening. Biomaterials 2010;31:5051e62. [76] Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined factors. Cell 2006;126: 663e76.