Actuation of Cell Sheets in 3D

TL;DR

Design principles used for programming curvature in nematic elastomers are applied to actuate 3D structures in the detached fibroblast layers, showing the ability to control the 3D shape through 2D patterning in cell layers, leading to promising avenues to program tissues.

Abstract

The alignment of fibers and cells in living tissues affect their mechanical properties and functionality. In this context, one can draw an analogy between tissues and nematic liquid crystal elastomers. We explore this analogy by growing fibroblasts on 2D-patterned substrates and observing the contraction of cell sheets upon detachment from the substrates. When fibroblast sheets detach, they undergo an anisotropic contraction, with maximum contraction along the nematic director, like nematic elastomers do during phase transition. We quantify this anisotropy using substrates patterned with stripes to induce alignment, finding that cell sheets resemble nematic elastomers with negative Poisson ratio. Then, we apply design principles used for programming curvature in nematic elastomers to actuate 3D structures in the detached fibroblast layers, demonstrating an application of these principles and we support the results with simulations. This proof of concept shows the ability to control the 3D shape through 2D patterning in cell layers, leading to promising avenues to program tissues.

Figures (9)

• Figure 1: Detachment of cell sheets from striped and unpatterned PDMS. (A) Phase contrast microscopy image of a cell sheet peeling off a PDMS slab with striped pattern (h=2$\mu$m, w=60$\mu$m). The bottom part of the image shows the substrate after cells have peeled off, while in the top part cells are still attached. Scale bar 200$\mu$m. (B) Image of an entire PDMS slab (outlined in lavender) after detachment is complete. The layer of cells, after peeling, is a smaller flat rectangle outlined in maroon dots. Scale bar 1000$\mu$m. (C) Schematic of the cell sheet peeled from PDMS with parallel stripes (top) and plain (bottom). For the striped pattern $x_i$ and $y_i$ are the initial dimensions of the PDMS sample parallel and perpendicular to the stripes, respectively. $x_f$ and $y_f$ are the final dimensions of the peeled sheet along the corresponding axes. For the plain PDMS $x_i$ and $y_i$ are defined as the dimensions of the longer and shorter axes of the PDMS, respectively, and $x_f$ and $y_f$ are the final dimensions of the peeled cell sheet along the corresponding axes. (D) Contraction of the cell sheet parallel and perpendicular to the alignment of the cells. Sample data from plain PDMS (here, $C_\parallel$ indicates the contraction along the long axis of the sample) are indicated in orange squares and and data for striped PDMS are indicated in purple triangles. The black dotted line corresponds to isotropic contraction ($C_\parallel=C_\perp$). Errors are described in Methods. The white square and triangle indicate the contraction for isotropic or aligned samples, respectively, obtained via simulation. (E) Final shape of simulated cell sheets with stripes (top) and without (bottom) and their height profiles. The color indicates the vertical deviation of the sheets with both remaining largely flat.
• Figure 2: Order parameter change upon detachment. (A-B) Fluorescence image of actin filaments and nuclei (stained with Phallodin for actic and NucBlue$^\mathrm{TM}$ Fixed Cell ReadyProbes$^\mathrm{TM}$ for nuclei, in insets) before (A) and after (B) the peeling of cells. Scale bars 100$\mu m$ in figure and 20$\mu m$ in insets. (C) Color maps of order parameter $S$ measured from images (A) on the left and (B) on the right after the analysis with OrientationJ (only half the images are shown). (C) Box plot of order parameter as defined in the paper, and calculated over 2 different samples (total n=21 images) before peeling and 2 different samples (total n=31 images) after detachment.
• Figure 3: Actuation of cell sheets. Schematic of the defect array patterns used in experiments (i), and a 3D rendering of the substrate topography with the height of the ridges at 1.5-2$\mu$m, ridge width 9$\mu$m (ii). (B) Confocal microscopy image of a sheet of cells detached from the defect array. The overall dimensions of the peeled sample are 2000$\mu$m x 2500$\mu$m. The color bar shows the height difference. XZ and YZ cross-sections along the directions indicated by white arrows reveal a sheet of uniform thickness with localized out-of-plane deformations. (C) Corresponding simulations of the cell sheet deformation. The inset on the bottom right corner represents a color map of Gaussian curvature (see SI for details). (D) Values of stretch factor $\lambda$ and Poisson ratio $\nu$ calculated from experiments (box plots for the flat samples with parallel ridges, and clear circles from curved samples with +1 topological defects) and simulations (filled rectangles for flat samples and clear rectangles for samples with +1 topological defects).
• Figure 4: Actuation of cell sheets with non-singular patterns. (A) Schematics of the tested alignment patterns featuring bend (left), and splay (right) patterns. All the patterns vary in 2-D and are made of short ridges like the schematic in \ref{['fig3']}A (ii) with the height of the ridges at 1.5-2$\mu$m, ridge width 9$\mu$m. (B-C) Confocal microscopy image of a peeled sheet of cells from the bend (B), and splay (C) patterns. The overall size of the detached samples are 1800$\mu$m x 3000$\mu$m (B), and 1800$\mu$m x 3000$\mu$m (C). The color bar shows the height difference. The cross sections show the consistent thickness of the sheet, and the curvature along the two principal directions for the bend (B) and splay (C) patterns, respectively. (D-E) Same as previous row but for simulated cell sheets, akin to what was seen in \ref{['fig3']}C. The insets in the corner are the maps of the Gaussian curvature of the sheets averaged over multiple simulations, see SI for details.
• Figure S1: Progression of peeling cell sheet over 4 minute period. (A) At t=0, or just before scraping the edges of the cell layer, cells are attached to the PDMS substrate, which is patterned with h=2$\mu$m stripe ridges with w=60$\mu$m spacing. The cells are aligned and elongated along the direction of the stripes. (B) At t=1min after scraping the edges, the cells are starting to peel from the substrate. The peeling front has reached the middle of the image. The fully lifted part of the cell sheet is visible near the bottom of the image (C) At t=2min, the peeling front has traveled to the top of the image. (D) After t=4min, the cells within the frame are almost fully peeled. Scale bars are 200$\mu$m.