Force exertion dynamics depend on substrate stiff-

Since p130Cas showed a clear mechanosensitive function in Cas WT cells, we hypothesized that its localization to FAs may also influence force ex- ertion itself. P130Cas is a known docking protein that influences FA dynamics on glass [12, 14], so we investigated the dynamics of force ex- ertion on micropillar arrays. Figure 4.4 shows still images from a time series movie of p130Cas-Venus YFP fluorescence in green and the tops of micropillars in red. As expected, random migration with an extended leading edge was visible after the cells were fully spread. Transient con- tractile forces opposite the direction of migration were visible at the leading edge of cells.

90 P130Cas in mechanosensing time (min) 0 2 4 6 8 10 11.6 kPa 137 kPa 47.2 kPa 20 nN Figure 4.4

Time-lapse imaging of force exertion and p130Cas. MEFs expressing p130Cas-Venus YFP (green) at endogenous levels were imaged every two minutes to examine the dynamics of force exertion on micropillars (red). Pillar deflections (white arrows) close to the automatically detected cell edge (blue) are plotted. Time stamp is in minutes above still images, force scale bar corresponds to 20 nN and image scale bar corresponds to 10µm (top-left image).

To test the effect of p130Cas localization to sites of force exertion, we quantified the force exertion characteristics of Cas -/- cells either re-expressing the full p130Cas-YFP (Cas WT) or a double-deletion mu- tant lacking both SH3 and CCH domains (∆SH3/∆CCH). This mutant has previously been characterized and was shown to lack FA localization while still being expressed at endogenous levels [14]. We confirmed that indeed, the truncated∆SH3/∆CCH version of p130Cas could not local- ize to FAs by fluorescence localization, whereas full-length p130Cas did localize to FAs on high-stiffness pillars as shown in Figure 4.3 and 4.4C. Cas ∆SH3/∆CCH cells still showed transient contractile forces during random migration.

Typical curves for force over time for single pillars are shown in figure 4.5A (blue dots correspond to data) for micropillar stiffnesses of 11.6 kPa, 47.2 kPa and 137 kPa, respectively. These transient force dynamics were modeled with a logistic function (see Methods for details) that is

4.3 Results 91

primarily described by two parameters: a maximum plateau force and a rate by which this maximum is obtained. The rate scales with the derivative of force over time but is independent of the plateau force. By fitting transient force exertion per pillar, we quantified the dynamics of local cellular force exertion. This model provided a good fit to the data, as shown in figure 4.5A (red line is the force increase fit and green line is the decrease fit).

The quantified plateau force (Fmax) and rate of force exertion (r)

are depicted in figures 4.5B and -C for all micropillar stiffness for both the Cas WT and Cas ∆SH3/∆CCH cells. Cas WT cells on micropillar arrays of stiffness 11.6, 47.2 and 137 kPa showed a mean force plateau of (12.8±0.5) nN, (23.5±0.9) nN and (39±2) nN, respectively (figure

4.5B). The ∆SH3/∆CCH mutant showed similar transient force exer- tion dynamics compared to the WT on 11.6 and 47.2 kPa micropillars. However, the force plateau was significantly higher for the WT in com- parison to Cas∆SH3/∆CCH on pillars with a stiffness of 137 kPa. The mean maximum force decreased to (27±4) nN for the Cas∆SH3/∆CCH cells.

There was also a significant difference in the rate of force exertion for Cas∆SH3/∆CCH compared to Cas WT cells. On micropillars with an effective Young’s modulus of 11.6 kPa and 47.2 kPa, there was no significant change in force exertion rate, for either Cas WT or Cas ∆SH3/∆CCH cells. However, the rate of force exertion for Cas WT cells on 137 kPa pillars significantly decreased from (0.75±0.05) min−1 _(on

47.2 kPa) to (0.45±0.03) min−1_{. In the absence of functional p130Cas}

localization to FAs (∆SH3/∆CCH mutant), the rate on 137 kPa pillars increased to (1.72±0.45) min−1_.

The observed dynamics of transient force exertion were thus slower when p130Cas was present for cells on high stiffness micropillars. On low stiffness micropillars we found no significant effect on force exertion with the localization of p130Cas to FAs. The presence of p130Cas in FAs (figure 4.3) thus directly correlated to changes in force exertion, but only when the global extracellular stiffness was larger than 47.2 kPa. The force became significantly larger when p130Cas was present in FAs and the dynamics were significantly faster when p130Cas did not localize to FAs.

92 P130Cas in mechanosensing

11.6 kPa 47.2 kPa 137 kPa

Time(min) F o rc e ( n N) 0 20 40 60 80 0 10 20 30 40 50 Time(min) F o rc e ( n N) 0 20 40 60 80 0 20 40 60 M a x F or ce ( n N)

Effective stiffness (kPa) p130Cas WT p130CasΔSH3/ΔCCH Ra te ( mi n -1)

Effective stiffness (kPa) 11.6 47.2 137 0 0.5 1 1.5 2 2.5 3 11.6 47.2 137 0 10 20 30 40 50 r Fmax dt dF ∝dF_dt B C A ** **** Time(min) F o rc e ( n N) 0 10 20 30 40 50 0 5 10 15 20 Increase fit Decrease fit Data Figure 4.5

Quantification of p130Cas-dependent force exertion dynamics. (A) Transient force exertion on a single pillar was quantified by fitting increase and decrease curves to a logistic function. From these fits the maximum force Fmax and the rate of

force exertionrwere obtained. Force curves were fitted for both Cas WT- and Cas

∆SH3/∆CCH MEFs. (B) In both cases, maximum force increased with increasing

stiffness. Cas ∆SH3/∆CCH MEFs exerted significantly less force on the high stiff-

ness arrays. Bar represent the mean Fmaxwith s.e.m. (C) The rate of force exertion

also shows a significant difference only on high-stiffness micropillars. (Significance by Kolmogorov-Smirnov test, ** p<0.01 and **** p<0.0001)