CONTROLLING INTERNAL ORGANIZATION AND DIVISION AXIS OF CULTURED CELLS WITH ADHESIVE MICROPATTERNS 1

M. Théry1, A. Pépin2, Y. Chen2, M. Bornens1

Institut Curie, CNRS UMR144, Paris, France, 2Laboratoire Photonique et Nanostructures, CNRS UPR20, Marcoussis, France

ABSTRACT We present a novel approach to control internal organization and polarity of individual adherent cells in culture. To control cell adhesion, we used micro-contact printing with two basic features: 1) the proper adhesive surfaces for a given convex envelope were reduced in order to impose more stringent adhesive conditions than a fully adhesive pattern with the same outline; 2) the respective positions of adhesive and non-adhesive zones in the convex envelope polarized the pattern. We show that cells, like in tissues, present a highly reproducible response when constrained by such stringent boundary conditions, as judged by the polarized organization of their internal compartments and by the orientation of their divisions. Keywords: micro-pattern, cell adhesion, polarity, division 1. INTRODUCTION Tissue differentiation during the development of multi-cellular organisms involves the organization of a great number of cells adopting a collective behavior. One of the basic features of cohesive multi-cellular organization, in which cell shape and migration are restricted by neighboring cells, is that each individual cells have the same polarity. Cell polarity is defined by the expression of a morphological and functional asymmetry relative to a polar axis defining a front and a rear edge. The actin cytoskeleton network is instrumental in maintaining structural and functional cell surface polarity, whereas the other major cytoskeleton network, the microtubule array, plays an essential role in controlling the asymmetric internal distribution of cytoplasmic organelles. The centrosome, which acts as the microtubule nucleating center, is physically associated with the nucleus and provides the microtubule astral network with an asymmetric distribution due to microtubules exclusion from the nucleus and a global polarity by anchoring the microtubules minus end. As a result, the intracellular compartments adopt an asymmetric and polar distribution. The symmetry axis defined by the nucleus, the centrosome and the Golgi apparatus thus reflects an internal distribution of compartments necessary for polarized intracellular transport and secretion of proteins, or for cell migration. This is particularly obvious in epithelial tissues where the plasma membrane itself is stably polarized, displaying two distinct regions: the apical domain facing the external environment or the lumen, and the baso-lateral domain, which is in contact with neighboring cells or a basal substratum (Figure 1). Cell polarity implies that cell surface polarity and internal polarity coincide and this is achieved by the formation of spatially and functionally distinct domains in the plasma membrane and the membrane–associated cytoskeleton. This organization must also be preserved during tissue growth. Cell divisions must be properly oriented in order to produce daughter cells able to insert in the tissue without impairing its cohesive integrity and function. Cell-cell junctions and cell adhesion to a basal substratum, which are connected to the intracellular cytoskeleton via transmembrane proteins, are clearly important features for that behavior.

Figure 1. Internal cell organization and polarity of an epithelial cell (left), a migrating cell (middle) and a cell on a [L] micro-pattern (right). The dashed arrow represents the polarized symmetry axis of the cell and the “+” and “-“ the microtubules polarity. 2. WORKING MODEL A basic question is whether polarity is an intrinsic and permanent property of cells, or if it is externally imposed by the tissue environment. Individual cells growing in culture like fibroblasts or keratocytes display a rather polarized migration (Figure 1) whereas other cell types show a more erratic behavior. In all cases, cells are permanently reshaping their cytoskeleton and their membrane activity. It is not easy to identify the basic principles of this (re)organization from the highly variable behavior of each individual cell. One possible way to experimentally overcome this riddle should be to grow individual cells in a fixed environment, for example on micro-adhesive patterns in which one single cell could spread and divide, but in which cell movement would be restricted like in a tissue. In these conditions, one could expect the inherent variability of individual cells to be normalized. If the boundary conditions imposed to cell adhesion possess a single and polarized symmetry axis due to the respective locations of adhesive and non-adhesive zones, it should drive the internal cell organization and polarity. As a result, the internal organization of each individual cells would be identical, like in tissues. But contrary to tissues, the internal organization of cells would become amenable to controlled modifications, and thus to the dissection of the interplay between intrinsic and externally imposed polarity. 3. EXPERIMENTAL To control the spatial distribution of the extra-cellular matrix we used micro-contact printing to manipulate the location of fibronectin [1]. Rather than using alkanethiolates selfassembled monolayers on gold we simply silanised glass coverslips with a mercapto-silane

after activation with a “piranha” solution [2]. We then printed the fibronectin (a tenth of which was labeled with Cy3) with a PDMS stamp and neutralized the non-printed areas with PEG-maleimide in order to make them repulsive to cell attachment. In addition to being cheap, easy to use and compatible with fluorescent cell-labeling, this technique allows a micrometer-resolution which is sufficient to design patterns at a subcellular scale. Thereby we imposed a similar shape to cells but distinct adhesion patterns (Figure 2). For comparison it was critical to impose identical cell spreading area instead of identical cell adhesion area [3]. HeLa cells, from a human adenocarcinoma epithelial cell line, and RPE1 cells, from a non transformed retinal pigment epithelial human cell line, were platted on micropatterns and fixed for staining or observed in time-lapse video-microscopy.

Figure 2. Living HeLa cells visualized on fibronectin-Cy3 micropatterns (a) with a 10x objective in phase-contrast (b). Cells display the same shape (triangle or square) on micropatterns having the same convex envelope. Bar = 10 µm. 4. RESULTS We observed and quantified the cell response to the spatial distribution of adhesion imposed by the micro-patterns. We first looked at the orientation of the internal cell polarity, defined by the respective location of the nucleus and the centrosome. It appeared that [L] micro-patterns guided the positions of the nucleus (close to the [L] corner) and the

Figure 3. Nucleus (gray ellipse on the left pictures), centrosome (two white dots on the left pictures) and Golgi (left) localization in fixed HeLa cells on [L]. Bar=10 µm. centrosome (close to the middle of the [L] hypotenuse) in a reproducible manner corresponding to the polarized symmetry axis of the [L] (i.e. their symmetry axis polarized by the respective location of the adhesive and non-adhesive border of the triangular

envelope) (Figure 3). The Golgi apparatus, rather than extended as it would be in classical culture conditions, was compacted around the centrosome. This reproducible structure allows to perform a refine quantitative study of intracellular traffic (Dimitrov A. et al, in preparation). Interestingly, cell polarity and centrosome positioning were more constrained on [L] than on full triangle (Figure 4). This showed that, in addition to cell shape, the restriction of the distribution of cell adhesion in a polarized manner directly impinges on cell polarity and contributes to reduce variability in cell internal organization.

Figure 4. Centrosome positions on [L] and on a full triangle were automatically detected and recorded in 200 fixed cells in each case. The disc area containing 90% of centrosomes (S90%) was 1,5 times bigger on [triangle] than on [L]. The non-adhesive area under the [L] hypotenuse also affected the architecture of the cell cytoskeleton. The bundle of actin stress fibers upon the non-adhesive border was more conspicuous than the one developed upon adhesive borders (Figure 5). It has already been shown that fully adhesive micropatterns confine cell protrusions and traction forces at cell apices [4]. On our micro-patterns the spatial distribution of adhesive and non-adhesive zones appeared to guide the assembly of the actin cytoskeleton and the directions and intensities of the traction forces the cell exert on the substrate, thereby influencing cell migration [5]. The microtubule network polarity also appeared guided by the [L] shape. Microtubules, nucleated from the centrosome, grew toward cell periphery where they followed actin cables mainly toward cell apices flanking the non-adhesive edge (Figure 5). Therefore, both cytoskeleton networks were influenced by the geometry and heterogeneity of the underlying pattern.

Figure 5. Actin cytoskeleton stained with phaloïdin and focal adhesions labeled with vinculin on full triangle and on [L] (left). Projection of a time lapse acquisition of microtubules “+” end trajectories in a living HeLa cell on a [L] (right). Bars = 10 µm.

Cell division is preceded by a bipolarization of the cell. This bipolarization emphasizes the polarity established in interphase : the two cell apices flanking the hypotenuse on [L] where large stress fibers were attached and where microtubules converged (Figure 5) defined the orientation of the spindle during division (Figure 6).

Figure 6. Metaphase cell showing from left to right chromosomes in phase contrast, mitotic spindle labeled with antibodies against tubulin, and centrosomes with centrin-GFP at spindle poles. The dashed line corresponds to the [L] symmetry axis, the solid line to the spindle axis. We plated cells on [L] and video-recorded their divisions in phase contrast microscopy. It revealed that they divided reproducibly along the hypotenuse (Figure 6a). An extensive comparison between the several micropatterns shown in figure 2 has been performed in order to dissect the respective role of cell shape and cell adhesion in the orientation of the division axis [6].

Figure 7. Three examples of RPE1 cells divisions illustrating the spatial and temporal reproducibility on [L] (a). [L] micropatterns also allow the control of two successive divisions of HeLa cells(b). The first division (3h3’) and the two following divisions (25h and 27h10’) were orientated along the direction of the [L] hypotenuse.

We were also able to record the second round of division of the two daughters cells. (Figure 6b). Interestingly the two daughter cells first respread on the two [L] branches (5h18’) and then reorganized into an apparently more stable configuration with one cell in the corner and the other along the hypotenuse (15h15’). The cell in the corner divided along [L] hypotenuse (25h) and its two daughters respread on the [L] branches. Then the second daughter of the first cell also divided along the hypotenuse (27h10’). After mixing together the four cells eventually found an apparently stable configuration respecting the symmetry conditions imposed by the underlying patterns (33h33’). These examples illustrated the ability of symmetrical and polarized micropatterns such as [L] to drive cell bipolarization during division. The two rounds of division reproduced how cells divide in a confined and polarized environment like in an embryo; this could open a new approach to investigate tissue morphogenesis. 5. CONCLUSION The reproducibility with which cells behave on the micro-patterns we designed (patent WO2005/026313) allows us to quantify the spatial distribution of cell compartments and of individual proteins and to assess its significance by averaging the distribution observed in a great number of cells. The ability of the micro-patterns symmetries and heterogeneities to impinge on cell internal organization and polarity up to mitosis could have a wide range of applications from the fundamental analysis of individual cell polarization or multicellular morphogenesis to tissue engineering. ACKNOWLEDGEMENTS We would like to thank Pierre Nassoy for help and advices on surface treatments and Donald E. Ingber and Philip LeDuc for generous gift of micro-patterns in the early part of our experiments. REFERENCES [1] Whitesides, G.M., E. Ostuni, S. Takayama, X. Jiang, and D.E. Ingber, Soft lithography in biology and biochemistry. Annu Rev Biomed Eng, 2001. 3: p. 33573. [2] Cuvelier, D., O. Rossier, P. Bassereau, and P. Nassoy, Micropatterned "adherent/repellent" glass surfaces for studying the spreading kinetics of individual red blood cells onto protein-decorated substrates. Eur Biophys J, 2003. 32(4): p. 342-54. [3] Chen, C.S., J.L. Alonso, E. Ostuni, G.M. Whitesides, and D.E. Ingber, Cell shape provides global control of focal adhesion assembly. Biochem Biophys Res Commun, 2003. 307(2): p. 355-61. [4] Parker, K.K., A.L. Brock, C. Brangwynne, R.J. Mannix, N. Wang, E. Ostuni, N.A. Geisse, J.C. Adams, G.M. Whitesides, and D.E. Ingber, Directional control of lamellipodia extension by constraining cell shape and orienting cell tractional forces. Faseb J, 2002. 16(10): p. 1195-204. [5] Jiang, X., D.A. Bruzewicz, A.P. Wong, M. Piel, and G.M. Whitesides, Directing cell migration with asymmetric micropatterns. Proc Natl Acad Sci U S A, 2005. 102(4): p. 975-8. [6] Théry, M., V. Racine, A. Pépin, M. Piel, Y. Chen, J. Sibarita, and M. Bornens, The spatial distribution of the extra-cellular matrix guides the orientation of the cell division axis. Submitted.