Supplemental Material

This article contains the following supporting material:

• Figure 1 - The effect of wild type Src and Fyn proteins on rigidity response in SYF cells was tested. Transfected cells were identified by immunoflorescent staining for activated SFKs. The effect of wild type constructs was similar to the effect of GFP-fusion proteins; reconstitution with Fyn increased the area ratio two-fold by suppressing spreading on soft substrates, while reconstitution with Src suppressed spreading regardless of substrate rigidity and the area ratio remained the same as in SYF cells (AR=1) (Supplemental Fig. 1A). The distribution of wild type Src and Fyn proteins was observed after 30 minutes of spreading on FN-coated glass. The immunofluorescence confirmed that wild type Fyn localizes to early focal contacts, while Src is distributed cytoplasmically (Supplemental Fig.1B). Since the localization of C-terminal GFP-fusion proteins was similar to wild type proteins, we further used only Fyn/Src-GFP. Scale bar: 20 μm.

• Figure 2 - To test whether proposed fibronectin rigidity response pathway is ligand-specific, we examined behavior of SYF, Cas-/-, and FAK-/- cell on collagen-coated polyacrylamide gels. The area ratios for SYF and Cas-/- cells were similar to control cells, while FAK-/- cells (previously identified as defective in spreading on collagen) have spread to larger area on soft than on rigid substrates (Supplemental Fig. 2A). In addition, morphology of FAK-/- cells on soft substrates appeared more similar to morphology of wild type fibroblasts. These results suggest that collagen rigidity response requires molecular machinery different than the rigidity response on fibronectin. Further, proliferation of SYF/Fyn-GFP, Cas-/-, and FAK-/- cells on collagen-coated gels was tested. Similar to control cells, growth of SYF/Fyn-GFP and Cas-/- cells was promoted by rigid, and not by soft collagen substrates. In contrast, FAK-/- cell proliferation was not affected by matrix rigidity (Supplemental Fig. 2B). Thus, we speculate that collagen rigidity response requires SFK-independent activation of FAK.

• Figure 3 - Spreading on FN-coated glass was quantified as percentage of cells spread after 20/30 minutes. While SYF cells transfected with the Fyn-GFP construct spread like control cells, introduction of Src-GFP failed to rescue the spreading phenotype to the same extent (Supplemental Fig. 3A). SRC+/+ cells (derived from Src+/+Fyn-/-Yes-/- littermates of SYF embryos) were used as control. In addition to differences in spreading 30 minutes after plating, a difference in the distribution of GFP-fusion proteins was observed. Similar to distribution of wild type proteins, Fyn-GFP was localized predominantly to the front of lamellipodia and in the tips of filopodial extensions, while Src-GFP fusion protein was uniformly distributed throughout the cytoplasm (Supplemental Fig. 3C). Transfection with the Fyn-GFP construct in RPTPα-/- background had less of an effect than in SYF cells (Supplemental Fig. 3B). In contrast to RPTPα+/+ (Supplemental Fig. 3D), the distribution of Fyn-GFP protein in RPTPα-/- fibroblasts was predominantly cytoplasmic (Supplemental Fig. 3E) suggesting that activation by RPTPα was necessary for the accumulation of Fyn at the leading edge. Scale bar: 10μm.

• Figure 4 - Role of RPTPα in FN-rigidity response was tested in MCF-10A human breast epithelial cells. The siRNA strategy previously used in fibroblasts was applied in MCF-10A cells. The FN-rigidity response was impaired in RPTPα-knockdown cells (Supplemental Fig. 4A) and the area ratio was decreased three-fold compared to area ratio in wild type MCF-10A (Supplemental Fig. 4A). The data is presented as mean+standard error for at least 50 cells. The individual knockdown cells were identified by immunofluorescence with an anti-RPTPα antibody (Supplemental Fig. 4B). Scale bar: 10μm.

• Figure 5 - Effect of palmitoylation site mutants on Cas phosphorylation was tested. SYF/Fyn(C3G)-GFP and SYF/Src(S3C)-GFP cells were plated on FN-coated glass for 45minutes and stained with an anti-phospho-Cas antibody. While the phosphorylated Cas localized to the focal contacts and leading edge in SYF/Src(S3C)-GFP cells, low levels of phosphorylated Cas were detected in SYF/Fyn(C3G)-GFP cells. Thus, Fyn-mediated Cas phosphorylation depends on Fyn palmitoylation and its consequent recruitment to the leading edge. Scale bar: 10μm.