Figure 1.
The initial cell spreading phase is stretch-independent while minimum stretching forces are necessary for cell orientation in the later phase of cell spreading.
(A) NIH3T3 fibroblasts were freshly seeded on fibronectin-coated membranes and cyclically stretched in uniaxial direction (double-headed arrow) with an amplitude of 8% at a frequency of 3 Hz. Cell spreading was monitored via time-lapse phase contrast microscopy. The cell contour of one exemplary cell is outlined in black. (Scale bar: 10 µm) (B) Kinetics of the mean cell adhesive area of initially non-adherent NIH3T3 fibroblasts at indicated stretch frequencies (control = non-stretched static condition). Time t = 0 indicates the time point at which cells were seeded onto the substrate. The data set can be divided into two groups depending on the applied frequencies (group I: low frequencies; group II: high frequencies) (ANOVA; group I compared to control: ‡p>0.05; group II compared to control: *p<0.05). (C) The mean cell orientation at indicated stretch frequencies over time. A mean value of 1 for the orientation parameter <cos2φ> indicates a perfectly-parallel, −1 a perfectly-perpendicular mean cell orientation with respect to the stretch axis. (ANOVA; group I compared to control: ‡p>0.05; group II compared to control: *p<0.05) (D) Cell elongation over time at indicated conditions. A value of 1 would describe a perfectly round cell, a value of 0 would be a perfect thin line. (ANOVA; group I and II compared to control: ‡p>0.05) (control, n = 89 cells; 0.1 Hz, n = 110 cells; 0.05 Hz, n = 80 cells; 1 Hz, n = 120 cells; 3 Hz, n = 117 cells; each from four independent experiments).
Figure 2.
Actin stress fibers reorganization accompanies the stretch-induced cell orientation of isotropic spreading cells.
(A) Time-lapse fluorescence imaging of a spreading NIH3T3 cell subjected to uniaxial CTS of a frequency of 3 Hz. The cell was double-transfected with Lifeact-GFP and mCherry-Vinculin. A cirumferential actin bundel system and seemingly thickend in parallel to the stretch axis were initially visible after cell seeding. Actin stress fibers realigned into a perpendicular orientation during the time-course of stretching. Cell-matrix adhesion sites emerged homogenously distributed along the cell edges independently of the stretch direction and reoriented during maturation into a perpendicular alignment during stretch application. (Scale bar: 10 µm) (B) Count of actin stress fibers confirm that in the initial phase of spreading mainly parallel transversal arc-like actin bundles were detected. The total number of actin stress fibers increased with time and perpendicular ventral actin stress fibers finally dominated. (C) Actin stress fiber orientation was analyzed under stretching conditions and static control (non-stretched) conditions. Left panel: The actin stress fibers oriented increasingly perpendicular to the stretch axis while they are not aligned under static conditions. Right panel: Actin fiber alignment was parallel to the major cell axis under stretch. Even though bipolar cell polarization occurs under control conditions the actin cytoskeleton showed no pronounced orientation with respect to the major cell axis. (T test, 36 min stretch compared to 36 min control: ‡p>0.05 for the right panel; *p<0.05 for the left panel. T test, 120 min stretch compared to 120 min control: *p<0.05 for the right and left panel.) (D) The orientation of focal adhesions in cells subjected to CTS and under control conditions was evaluated. Focal adhesions oriented perpendicular to the stretch axis (left panel) and thus parallel to the major axis of the perpendicular oriented fibroblasts (right panel). Under non-stretched control conditions no significant preferential alignment of focal adhesion with respect to an arbitrary x-axis or the major cell axis was observed. (E) The number and orientation of protrusions was determined under stretch conditions for Life-Act-GFP and mCherry-Vinculin expressing cells. Protrusions that formed at the sides of the cells were defined as parallel; protrusions formed at the end of the cell were assigned as perpendicular to the stretch axis. The number of protrusion was at the beginning equally distributed around a spreading cell with a slightly preferred parallel formation. With increasing time of exposure to CTS the protrusive activity was dominantly at the perpendicular ends of the cell. (T test, side compared to end: ‡p>0.05 for 10 min and 16 min; *p<0.05 for 32 min and 46 min) (Stretched condition, n = 21 cells; non-stretched static control conditions, n = 11).
Figure 3.
The stretch-induced perpendicular and polarized cell spreading is microtubule-independent but myosin II-dependent.
(A) Spreading NIH3T3 fibroblasts subjected to uniaxial CTS of 8% amplitude at 3 Hz (double-headed arrow indicates the stretch direction). The cell was pre-treated with blebbistatin. The cell outline of one exemplary blebbistatin-treated cell is given. (Scale bar: 10 µm) (B) Kinetics of the mean cell adhesive area of spreading cells at 3 Hz at indicated conditions. The time point t = 0 indicates when the pharmacologically pre-treated cells were seeded onto the substrate. (ANOVA, pharmacologically treated cells compared to non-treated cells: *p<0.05) (C) Kinetics of the stretch-induced mean cell orientation of cells treated at 3 Hz with pharmacological substances. Blebbistatin-treated cells tend to orient slightly parallel to the stretch axis while microtubule-disturbed cells aligned perpendicular to the stretch direction. (ANOVA, nocodazole- and taxol-treated cells compared to non-treated cells: *p<0.05; blebbistatin-treated cells compared to non-treated cells: ‡p>0.05) (D) Time-course of the cell elongation at 3 Hz at indicated conditions. (ANOVA, pharmacologically treated cells compared to non-treated cells: *p<0.05). (nocodazole = disrupts microtubules; taxol = stabilizes microtubules; blebbistatin = inhibits myosin II activity) (Non-treated, n = 117 cells; Blebbistatin, n = 83 cells; Nocodazole, n = 104 cells; Taxol, n = 95 cells; each from four (non-treated sample) to three (pharmacological-treated samples) independent experiments).
Figure 4.
Overview of cell spreading and orientation under cyclic stretch conditions.
Spreading of fibroblasts upon CTS application occurs in two phases. In the first phase (“Spreading”) the initial cell attachment is generated. Then a circular lamellipodia and dot-like cell-matrix adhesion sites are visible. The cell reaches its critical adhesive area. In the beginning of the second phase (“Polarization/Orientation”) the cell adhesive area reaches its maximum and cell elongation is initiated. The cell develops with increasing time of stretching pronounced actin stress fibers and cell-matrix adhesions which become with time perpendicularly oriented with respect to the stretch axis. In the spreading phase the formation of actin bundles in parallel to the stretch axis is observed while cell-matrix adhesions emerged homogenously distributed along the cell edges independently of the stretch direction. Cell-matrix adhesions sites reoriented into a perpendicular alignment and the parallel actin fibers realign into a perpendicular orientation and partially disassemble as the stretching force continues to act on the cell.