The stretch effect

If you tread on someone's toe, the response is usually quite vocal. But, beyond the yelps, what might be happening when force is applied at the cellular level? Tamada, Sheetz and Sawada investigated how forces on the extracellular matrix (ECM) are transduced into intracellular biochemical signals, using cells that have been stripped of their cell membrane and soluble proteins. Stretching the resultant 'Triton cytoskeletons' initiated signalling to the small GTPase Rap1, as outlined in Developmental Cell.

When cells were extracted using Triton X-100, the remaining complex contained mainly cytoskeletal and adhesion proteins, and only a few membrane lipids or cytoplasmic proteins. Nevertheless, the authors found that, when these Triton cytoskeletons were stretched, they could activate Rap1 that was present in added cytoplasmic extracts. Such stretch-mediated Rap1 activation is known to occur in intact cells.

On investigating potential upstream candidates, the authors found that two components of an added soluble cytoplasmic extract, the Rap1 guanine nucleotide-exchange factor C3G and the adaptor protein CrkII — which are both involved in activating Rap1 — bound to the Triton cytoskeletons in a stretch-dependent manner. Activation of Rap1 was prevented if C3G was depleted from the extracts, so the Rap1 response to Triton-cytoskeleton stretching depended on C3G.

In response to various stimuli, CrkII–C3G binds to tyrosine-phosphorylated proteins through the Src-homology-2 (SH2) domain of CrkII. And, accordingly, Tamada, Sheetz and Sawada noticed an increase in phosphotyrosine levels in several protein components of Triton cytoskeletons in response to stretch. The prime suspects for this phosphorylation were Src-family kinases (SFKs), and a selective SFK inhibitor prevented this stretch-induced phosphorylation response and the ability of fluorescently tagged CrkII to bind to stretched Triton cytoskeletons. A candidate substrate for such phosphorylation was Cas (Crk-associated substrate), a significant amount of which remained on Triton cytoskeletons. The authors' assumptions proved correct — tyrosine phosphorylation of Cas in Triton cytoskeletons increased in response to stretch. And CrkII no longer bound to Triton cytoskeletons of Cas^−/− cells, which indicates that CrkII binds directly to tyrosine-phosphorylated Cas.

Does this affect all parts of the cell in the same way? The greatest stresses are expected to occur at the regions of cell–ECM contact. At these points, Tamada, Sheetz and Sawada observed an increase in tyrosine phosphorylation in Triton cytoskeletons in response to externally applied force — no cytosolic molecules were required. A similar localized increase in tyrosine phosphorylation was seen when intact, non-Triton-treated cells were stretched. And fluorescently tagged CrkII was seen to move to adhesion sites and to colocalize with Cas in response to stretching in intact cells.

So stretching Triton cytoskeletons and intact cells induces CrkII–C3G–Rap1 signalling. How the signalling is initiated is unknown, but it's likely that proteins are unfolded or distorted when force is applied, which could create new binding sites for other proteins. As Rap1 is known to induce integrin-mediated adhesion, its activation at cell–ECM sites could well stabilize such contacts. Recent insights into the activation of integrins by Rap1 come from the studies of Lafuente et al., who cloned and characterized RIAM, a Rap1–GTP-interacting adaptor molecule. When overexpressed, RIAM induced β₁- and β₂-integrin-mediated adhesion and influenced actin dynamics.