Figure 1.
Interplay of elongation and contraction in stretching chambers.
(A) An incompressible elastic material in uniaxial stress (i.e. forces act exclusively along one direction) exhibits a transversal shrinkage of exactly half the elongation along the direction of force. That is, upon elongation of an elastomeric film in x-direction, a film characterized by a Poisson's ratio of 0.5 is shortened in y-direction with Δy = ½ Δx. For adhered cells this results in tensile strain in x- and contractile strain in y-direction. (B) Scheme of the stretching design. An elastomeric chamber was formed of silicone rubber exhibiting a bulky 5 mm thick wall all around an approximately 400 µm thin film (Young's modulus 50 kPa, Poisson's ratio = 0.5). The chamber (length 20 mm, width 20 mm) was clamped on two opposite sides and stretched with controlled amplitudes A in x-direction. Exact substrate deformations applied to the cells were quantified by mapping a regular micro-pattern embedded in the silicone rubber film. All cells adhered in the indicated cross were analyzed. Ribbon-like chambers display an identical design except that bulky walls in x-direction (direction of stretch) are missing. (C) Dependence of zero strain direction on the measured transversal shrinkage factor κ (κ = −Δy/Δx). In theory κ can assume values from 0 (fully blocked perpendicular shrinkage upon film elongation) to 0.5 (free shrinkage upon elongation).
Figure 2.
Cell reorientation upon cyclic stretch.
Stretch cycles as indicated (top left) of various amplitudes (a0 (0%), a1 (4.9%), a2 (8.4%), a3 (11.8%), a4 (14.0%), a5 (32%), b5 (31.7%)) were applied to adhered cells for 16 hours. Subsequently, cells were fixed and stained for actin. The arrow indicates the direction of stretch. Note the induced cytoskeletal reorientation upon cyclic stretch. Frequencies used for each amplitude are given in table 1. Scale bars, 50 µm.
Figure 3.
Determination of cell shape and cytoskeletal orientation.
(A) Cells were stained for actin after stretch application and cell fixation. Every cell and its actin bundles were recognized by image processing. Actin bundle directions were determined at every single pixel as perpendicular direction to the grey value gradient (inlay). The major axis of the ellipse with the same normalized second central moments as the cell characterized the main cell orientation. Black arrows next to the cells indicate main actin orientation, and the white lines give the cell shape orientation. (B) Histogram of measured actin orientations of the cells 1 to 3 given in A. All values of each cell were used to determine the main actin orientation angle indicated for the three cells (diamonds). See also Table 3 or Figure 4.
Table 1.
Average cell aspect ratios.
Table 2.
Mean cell orientations.
Figure 4.
Orientation of stress fibers in response to uniaxial cyclic stretch.
The predominant cytoskeletal orientations were determined for cells stretched for 16 hours at amplitudes a1 (4.9%, 1493 cells), a2 (8.4%, 662 cells), a3 (11.8%, 625 cells), a4 (14.0%, 397 cells), and a5 (32%, 772 cells) in box-shaped chambers (κ = 0.15) and at amplitude b5 (31.7%, 588 cells) on ribbon-like substrates (κ = 0.29). Control measurements in both types of chambers (a0 (0%, 1417 cells), b0 (0%, 640 cells, b0 only indicated as cumulative plot)) showed random distributions. Distributions are given as histograms and cumulative distributions (bottom right). The dashed line in each histogram indicates a random distribution. The dotted lines express the theoretical description of the measured actin angle distributions (see results section for details). All distributions differ significantly from each other according to Kolmogorov/Smirnov except a0 from a1 and a3 trom a4.
Table 3.
Distribution of actin angles.
Figure 5.
Time evolution of cytoskeletal (A, actin) and cell orientation (B, whole cell) upon cyclic strain of 11.8% amplitude (a3).
Control data (0 min) were taken on cells cultivated for 5 hours without applied strain. The final reorientation distribution in A and B is indicated by the dashed black line. Cell numbers in order of increasing duration: 72, 80, 227, 187, 159, and 261 cells. Note that for (A) all time points from 180 min and higher differ significantly from control while in (B) all distributions from 30 to 300 min equal random distributions (p = 0.05 according to Kolmogorov-Smirnov test). The control distribution for 0 min is slightly shifted due to the relatively low number of cells evaluated for this time point. In (C) exemplary micrographs of cells stretched with amplitude a3 for indicated times and subsequently labeled for actin are shown. Staining was done in parallel and microscope settings were kept identical to allow comparison of labeling intensities.
Figure 6.
Actin and vinculin concentrations increase in stretched cells.
Cells were cyclically stretched with amplitudes a2 (8.4%) and a4 (14%), respectively, and immunofluorescently labeled for actin and vinculin. Mean grey values for actin (dark grey bars) and vinculin (light grey bars) were determined and are given relative to the control (a0) values. For each stretching experiment actin bundles from 20 cells and 100 FAs from 5 to 7 cells were evaluated. Asterisk = significantly different to a0 (P≤0.05).
Figure 7.
Tyrosine phosphorylation during stretch application.
Cells were cyclically stretched with amplitude a3 (11.8%) and immunofluorescently stained for phospho-tyrosine after indicated durations. All time points were stained simultaneously. Mean grey values are given. n = 50 focal adhesions from 5 to 6 different cells. Asterisk = significantly different to initial values (0 h) (P≤0.05). Typical images analyzed for phospho-tyrosine staining of FAs are indicated below the graph. Here, microscope settings were kept identical for all images and show therefore basically no phosphorylation at time points 0 h and 16 h.
Figure 8.
Effect of substrate elasticity on mechanoresponse induction.
Cells on stretch chambers of ∼1 kPa, 3 kPa, 11 kPa and 50 kPa elasticity were stretched for 16 hours with amplitude a3 (11.8%). Subsequently, cells were fixed, stained for actin and main actin orientation angles (A) as well as main cell orientations (B) were determined. n equals 145, 86, 152 and 261 cells for Young's moduli of ∼1 kPa, 3 kPa, 11 kPa and 50 kPa, respectively. (C) Single cell analysis on cells grown on ∼1 kPa and 50 kPa substrates, respectively, before and after the indicated substrate stretch (Y, black arrow). Substrate elongations under the cell (X, white arrow) were determined. The indicated percentage is the effective substrate elongation. Note the strongly reduced substrate stretch under the cell on ∼1 kPa substrates. n = 2 cells on ∼1 kPa substrate and 3 cells on 50 kPa substrate.
Table 4.
Influence of substrate stiffness on cell property during stretch.