Langmuir 2009, 25, 367-376
367
Identification of Three Regimes of Behavior for Cell Attachment on Topographically Patterned Substrates Peter M. Stevenson and Athene M. Donald* Biological and Soft Systems, CaVendish Laboratory, Department of Physics, UniVersity of Cambridge, JJ Thomson AVenue, Cambridge CB3 0HE, U.K. ReceiVed September 1, 2008. ReVised Manuscript ReceiVed October 14, 2008 3T3 fibroblasts were cultured on poly(dimethyl siloxane) line-shaped ridge/groove topography with a range of ridge widths (25-55 µm), ridge spacings (10-80 µm), and two ridge heights (15 and 21 µm). Three distinct regimes of attachment occurred, which were dependent on the ridge spacings used. Using the 21 µm height ridges, at the smallest spacings (∼10-20 µm) cells were able to bridge between neighboring ridges without touching the groove floor below. At moderate spacings (∼30-50 µm), cells were confined to a single ridge or groove only and aligned to the pattern at an increasing degree as the ridge width narrowed. The largest spacings (>∼50 µm) allowed cells to connect between a ridge and a groove, and the connection occurred most frequently at angles nearly perpendicular to the pattern. Reducing the ridge height to 15 µm allowed ridge-groove connections also at 40 µm spacings but had no effect on bridging or alignment. It was proposed that both a critical length and a critical angle (slope) exist for any cell protuberance that connects between a ridge-ridge or ridge-groove. These results build on previous studies by using a single cell type and focusing quantitatively on the regimes permitting different modes of spreading. In addition, particular focus on the ridge-groove connections has allowed more comprehensive quantification of the incident angles and morphologies of cells as they connect between a ridge and a groove.
Introduction The use of surface modification allowing controlled, patterned culture of mammalian cells is a valuable tool in cellular assays. In the body, cells organize in and assemble in defined patterns to create functional tissues such as with oriented capillary growth or extension of neural networks.1 Homogenous laboratory culture vessels do not contain the complex 3D cues that can instruct cells to differentiate, proliferate, or orient in a particular conformation, which is needed to form complete tissues. Although it is difficult to mimic this environment in vitro, some progress can be made by using simple topographical or immobilized chemical modifications to impose some local changes in the cellular response. Depending on the cell type, simple shape changes such as elongation can improve adhesion strength,2 correct differentiation/function,3,4 and migration speeds.5,6 With respect to the latter, it has been noted that cells align in the direction of migration,7 and thus forcing cells to orient in a particular direction may allow direct control over cell migration. The spatial control of cells has also allowed coculture of different cell types on a single surface, and various other useful functionalities with respect to biomaterials.1 Chemical patterns often include surface modifications that inhibit attachment in certain areas, and thus cells are restricted not only in the way they orient but also in outright attachment, which is not ideal for tissue engineering applications.1 * To whom correspondence should be addressed. E-mail:
[email protected]. (1) Kumar, G.; Ho, C.-C.; Co, C. C. AdV. Mater. 2007, 19, 1084–1090. (2) Eisenbarth, E.; Linez, P.; Biehl, V.; Velten, D.; Breme, J.; Hildebrand, H. F. Biomolecular Engineering 2002, 19, 233–237. (3) Lu, L.; Kam, L. C.; Hasenbein, M.; Bizios, R.; Mikos, A. G. In Human Retinal Pigment Epithelium Cell Culture On Patterned Surfaces; Department of Biomedical Engineering, Rensselaer Polytechnic Institute: Troy, NY. (4) Recknor, J. B.; Recknor, J. C.; Sakaguchi, D. S.; Mallapragad, S. K. Biomaterials 2004, 25, 2753–2767. (5) Li, S.; Bhatia, S.; Hu, Y.-L.; Shiu, Y.-T.; Li, Y.-S.; Usami, S.; Chien, S. Biorheology 2001, 38, 101–108. (6) Rajnicek, A. M.; Britland, S.; McCaig, C. D. J. Cell Sci. 1997, 110, 2905– 2913. (7) Boocock, C. A. DeVelopment 1989, 107, 881–890.
Topographical patterns may provide more insight into cell patterning because chemistry can be maintained across the substrate (assuming the topography does not affect aspects such as wettability), thus removing the choice for adhesive regions over repellent ones. A simple but effective geometry that has been used in previous studies involves line-shaped features with repeating ridges and grooves.4,8-20 Several of these studies have demonstrated that cells do indeed elongate and align their extended morphology with the direction of the lines. For example, Clark et al.18 tested fibroblast adhesion on 4-24 µm width repeating ridges and grooves and observed that the cells aligned almost entirely with the pattern when the height of the ridges was 1.9 µm (the maximum tested height). Reduction in the depth resulted in progressively diminished alignment as cells became able to span across several ridges, thus causing a deviation from alignment. The critical groove depth (ridge height) needed to impart alignment at any given ridge/groove scale has been (8) Chou, L.; Firth, J. D.; Uitto, V.-J.; Brunette, D. M. J. Cell Sci. 1995, 108, 1563–1573. (9) van Kooten, T. G.; Whitesides, J. F.; von Recum, A. F. J. Biomed. Mater. Res. 1998, 43, 1–14. (10) Mrksich, M.; Chen, C. S.; Xia, Y.; Dike, L. E.; Ingber, D. E.; Whitesides, G. M. Proc. Natl. Acad. Sci. U.S.A. 1996, 93, 10775–10778. (11) Teixeira, A. I.; Abrams, G. A.; Bertics, P. J.; Murphy, C. J.; Nealey, P. F. J. Cell Sci. 2003, 116(10), 1881–1892. (12) den Braber, E. T.; de Ruijter, J. E.; Ginsel, L. A.; von Recum, A. F.; Jansen, J. A. J. Biomed. Mater. Res. 1998, 40, 291–300. (13) Curtis, A. S. G.; Casey, B.; Gallagher, J. O.; Pasqui, D.; Wood, M. A.; Wilkinson, C. D. W. Biophys. Chem. 2001, 94, 275–283. (14) Clark, P.; Connolly, P.; Curtis, A. S. G.; Dow, J. A. T.; Wilkinson, C. D. W. J. Cell Sci. 1991, 99, 73–77. (15) Dunn, G. A.; Brown, A. F. J. Cell Sci. 1986, 83, 313–340. (16) Goldner, J. S.; Bruder, J. M.; Li, G.; Gazzola, D.; Hoffman-Kim, D. Biomaterials 2006, 27, 460–472. (17) Sarkar, S.; Dadhania, M.; Rourke, P.; Desai, T. A.; Wong, J. Y. Acta Biomaterialia 2005, 1, 93–100. (18) Clark, P.; Connolly, P.; Curtis, A. S. G.; Dow, J. A. T.; Wilkinson, C. D. W. DeVelopment 1990, 108, 635–644. (19) Loesberg, W. A.; te Riet, J.; van Delft, F. C. M. J. M.; Scho¨n, P.; Figdor, C. G.; Spellerd, S.; van Loon, J. J. W. A.; Walboomers, X. F.; Jansen, J. A. Biomaterials 2007, 28, 3944–3951. (20) Alaerts, J. A.; Cupere, V. M. D; Moser, S.; Van Den Bosh de Aguilar, P.; Rouxhet, P. G. Biomaterials 2001, 22(2), 1635–1642.
10.1021/la802859v CCC: $40.75 2009 American Chemical Society Published on Web 12/09/2008
368 Langmuir, Vol. 25, No. 1, 2009
identified as an important factor elsewhere - for example 0.5 µm deep polystyrene grooves exhibited significant fibroblast alignment at 1 µm groove/ridge width but not at 2, 10, and 20 µm widths.21 In fact, fibroblasts have been shown to align on depths as small as 35 nm when using 100 nm width lines.19 As such, it is clear there is a relationship between the height/depth and width of ridges/grooves respectively needed to induce alignment, although the scales tested vary so much that it is difficult to compare studies quantitatively. Once cells are highly aligned on the tops of ridges, it has been shown that further substantial increase in ridge height has little if any effect on alignment (in the 10s of microns height scale at least22), whereas there is a proportional response between reducing ridge width and increased alignment.15,18 The highly aligned cells in Clark et al.’s18 test case had cell bodies that rested on several ridges, but they still had near-bipolar elongations in the direction of alignment along the ridges. This coexistence of alignment while spanning several ridges has been shown elsewhere for fibroblasts in the low-micron scale (∼10 µm).8,12,20,21 Therefore, alignment does not necessarily have to be induced on a single ridge or groove only; however, conversely the term “bridging” refers to a more specific case where cells span neighboring features with a coinciding breakdown in alignment. The maximum distance of cell bridging depends on the cell type, but fibroblasts can readily bridge 10 µm,23 whereas neurons have been shown to bridge gaps of up to 100 µm,16 with increasing frequency as the gap size is reduced. In ridge-ridge bridging, cells are assumed not to touch the groove floor below,23 and thus the height and angle of the topographic step at the edge of a ridge and groove is considered to be unimportant. However, both of these factors can determine whether cells can connect between ridges and grooves. Whereas several studies have presented images where cells can be seen to span from a ridge to a groove, on several occasions these events are not actually addressed in any great detail because they are focusing on other aspects of cell patterning.21,22 Often, the primary focus in any discussion of cells overcoming a topographic boundary is the bending of the cytoskeleton to facilitate it,15,18,24-26 but there are many variations and limitations to the observed results, which make comparisons difficult. Clark et al.24 investigated single, isolated 1-18 µm height steps and found that fibroblasts were able to connect from the top of the step to the floor below (or vice versa) but with decreasing frequency as the step height increased. In addition, for heights of 10 µm or more the cells appeared to climb down the wall of the step, thought to be due to difficulty of sustaining a suspensory cell protrusion from the top to the floor without at least some substrate contact. Cells that did not connect between the step and floor aligned along the boundary, with increasing frequency as the step height increased. Dunn and Heath25 used prism-shaped ridges with varying angles and found that angles >4° restricted cell locomotion at the peak boundary, with cells often aligning along the boundary rather than crossing it. Cells that did cross the boundary often had microfilaments that were isolated on either side of the boundary or they crossed at angles tending to lie parallel to the boundary. Singular topographic steps such as these allow a simpler investigation of a cellular reaction to the boundary but do not have to factor the space available in the grooves or other influences (21) Walboomers, X. F.; Monaghan, W.; Curtis, A. S. G.; Jansen, J. A. J. Biomed. Mater. Res. 1999, 46, 212–220. (22) Brunette, D. M. Exp. Cell Res. 1986, 164(1), 11–26. (23) Ohara, P. T.; Buck, R. C. Exp. Cell Res. 1979, 121, 235–249. (24) Clark, P.; Connolly, P.; Curtis, A. S. G.; Dow, J. A. T.; Wilkinson, C. D. W. DeVelopment 1987, 99, 439–448. (25) Dunn, G. A.; Heath, J. P. Exp. Cell Res. 1976, 101, 1–14. (26) Brunette, D. M. Exp. Cell Res. 1986, 167, 203–217.
SteVenson and Donald
of neighboring features when using repeating patterns. When using repeating features, the choice of 90° steps16,18,22 or more subtle topographic boundaries (V-shaped grooves,22,26 rounded humps27) leads to differences in how much the cytoskeletal filaments will need to distort. Furthermore, in several studies the groove depths are very small (
