Fig 1.
Representative example of the process for dual nuclear and Golgi tracking.
Image tracking. Images of nuclei (blue) and Golgi bodies (green) are shown before (A,C) and after (B,D) activation, where the wrinkle direction is vertical in all images shown. Nuclei were fit with ellipses (shown in red) before (A) and after (B) wrinkling, with orientation defined as the ellipse major axis. In addition, we identified both the nucleus and Golgi body to determine cell orientation before (C) and after (D) wrinkling, indicated by red arrows between a Golgi body and nucleus in the same cell. Polar histograms in the bottom right show a similar mean and truncated standard deviation. Scale bar is 100 μm.
Fig 2.
Topographic change of the shape memory polymer substrates.
Optical micrographs of an active wrinkling shape memory polymer substrate (A) before and (B) after completion of formation of the nano-wrinkled topography employed to examine the temporal evolution of mouse fibroblast polarization. The direction of programmed tensile strain in (A) is horizontal, and, thus, the direction of wrinkle formation in (B) is vertical, as indicated by the double-headed arrow. Scale bar is 50 μm.
Fig 3.
(A-I) Mouse fibroblast cell trajectories with a normalized origin on wrinkled (W; A-C), non-wrinkled (NW; D-F) and active (A; G-I) substrates for each time region. Trajectories have been rotated so that the x-axis corresponds with the wrinkle direction. Cells on static non-wrinkled substrates explore space equally in all directions, while cells on static wrinkled substrates show preferred motion along the wrinkle direction. (J) Ratio of the components of the radius of gyration tensor along (R_xx) and perpendicular to (R_yy) the direction of wrinkles, quantifying the degree of anisotropy in cell trajectories on non-wrinkled (blue), wrinkled (green) and active (red) substrates. Cells on non-wrinkled substrates explore space equally in all directions, while cells on wrinkled and active substrates prefer motion along substrate wrinkles. (K) The ratio of the x and y components of mouse fibroblast velocities also quantifies the anisotropy of cell trajectories, averaged over all times. Cells on wrinkled and active substrates have significantly higher speeds along wrinkles, while cells on non-wrinkled substrates have approximately equal speeds in all directions. Single asterisks (*) indicate significance levels below 0.05, while double asterisks (**) indicate levels below 0.01. There were approximately 103 cells per substrate type across 3 technical replicates and 3 biological replicates.
Fig 4.
Mouse fibroblast velocity auto-correlation function.
Velocity auto-correlation function (VACF) of control (solid lines) and ROCK-inhibited (dotted lines) mouse fibroblast cells on non-wrinkled (NW), wrinkled (W), and active (A) substrates. Given the proximity of each curve, we have excluded error bars in this figure so that more data would be visible. Curves are cut off at a timescale of 10 hours, after which the signal to noise ratio is low enough that results are unreliable. The black line shows a best-fit slope for all of the VACFs with an exponential decay of approximately 4 hours, corresponding to the transition timescale between ballistic and diffusive motion for trajectories in these systems. It is remarkable that even when the ROCK pathway is inhibited cells retain roughly the same diffusion timescale, indicating that ROCK-inhibition is not directly interfering with persistent cell motion, but rather the ability of cells to sense and align with wrinkles, quantified by the truncated standard deviation (TSD).
Fig 5.
Control mouse fibroblast orientation statistics.
Mean cell orientation with respect to the wrinkle direction, in degrees, (A,C) and truncated standard deviation of orientations, in degrees, (B,D) for nuclei orientation (A,B) and the axis between the nuclei and Golgi body centers-of-mass (NGV) (C,D) for control mouse fibroblast cells on non-wrinkled (NW), wrinkled (W), and active (A) substrates. Importantly, the truncated standard deviation (TSD) of nuclei orientation decreases over time, showing increasing alignment. Furthermore, the NGV has a truncated standard deviation similar to wrinkled substrates at all times, indicating that this definition of orientation could be more sensitive to environmental cues, whether pre-programmed patterns or substrate topography. (E) To elucidate the time evolution of alignment, we show a times series for nuclei truncated standard deviation, in degrees, fit with a decaying exponential (black line) with a timescale of approximately 2 hours. (F) Time evolution of NGV truncated standard deviation, in degrees. Single asterisks (*) indicate significance levels below 0.05, while double asterisks (**) indicate levels below 0.01. There were approximately 103 cells per substrate type across 3 technical replicates and 3 biological replicates.
Fig 6.
ROCK-inhibited comparisons with control mouse fibroblasts.
A comparison between velocity ratio (A), mean nuclei orientation, in degrees, (B), and NGV truncated standard deviation, in degrees, for control (solid) and ROCK-inhibited (lined) systems. Inhibiting the ROCK pathway does not prevent those systems from reaching a state statistically similar to control wrinkled substrates. However, ROCK inhibition has a large effect on active substrates where it prolongs or prevents reorientation. Single asterisks (*) indicate significance levels below 0.05, while double asterisks (**) indicate levels below 0.01. There were approximately 103 cells per substrate type across 3 technical replicates and 3 biological replicates.