Figure 5. Force-dependent binding of cadherin/catenin complexes to F-actin.

(A) Mean binding lifetimes (red filled circles) from constant-force assay measurements from previously reported (Bax et al., 2022) wild-type E-cadherin/β-catenin/αE-catenin complex data (N = 700). These data are represented here without depicting the direction of force applied relative to the polar actin filament, and fit to a nondirectional two-state catch bond (red curve). Unfilled circles represent the mean lifetime of events collected in the low-force assay (N = 90). Envelopes indicate 95% confidence intervals for the fit, obtained by empirical bootstrapping. Areas of all circles are proportional to the number of events measured in each equal-width bin. (B) Mean binding lifetimes (blue filled circles) from pooled low- (N = 145) and constant-force assay (N = 856) measurements for ternaryΔH1 complexes. These data were fit to a one-state slip bond model (blue curve). (C) The one-state slip bond model. The conformation of a bound actin-binding domain (ABD) missing H1, denoted state B, is comparable to the strong state of the two-state catch bond model. Molecules transition between bound (B) and unbound (U) states, where the dissociation rate, ${\displaystyle k_{\mathrm{B}\to \mathrm{U}}}$ , increases exponentially with force.

Figure 5—figure supplement 1. Two-state slip bond model.

In this model, the ternaryΔH1 complex interacts with F-actin in a moderately strong bound state (B₁) and strongly bound state (B₂), such that that the deletion of H0 and H1 hinders the transition between B₁ and B₂ to impart biphasic binding lifetimes. However, Akaike information criterion (AIC) and Bayesian information criterion (BIC) suggest a single-state slip bond model better represents the data (see text).

Figure 5—figure supplement 2. Lifetime survival analysis for wild-type E-cadherin/β-catenin/αE-catenin and E-cadherin/β-catenin/αE-cateninΔH1.

Survival probability versus binding lifetime from constant-force measurements for each 2 pN bin width. (A) Survival lifetimes are more likely to be longer for ternaryΔH1 (blue) than wild-type ternary (red) between 0 and 2 pN. (B) The ternaryΔH1 complex survival probability distribution is qualitatively more monophasic when compared to wild-type ternary. (C) Binding lifetimes for ternaryΔH1 are more likely to be shorter than wild-type ternary complex between 4 and 6 pN. The ternaryΔH1 complex is qualitatively more monophasic than wild type. (D–F) Above 6 pN, ternary wild-type and ternaryΔH1 complex survival probability distributions at these force bins are not statistically different (p < 0.01, two-sample Kolmogorov–Smirnov [KS] test).

Figure 5—figure supplement 3. Molecular basis of catch bond directionality.

(A) Force on actin (gray arrow) directed in the (−) direction of ternary wild type. When force is applied in the (−) direction, H1 is moved away from H2–H5 and oriented such that reassociation of H0/H1 to the H2–H5 bundle is disfavored (dotted lines). (B) Force on actin (gray arrow) directed in (+) direction of ternary wild type. When force is applied in the (+) direction, H1 is predicted to be positioned relatively closer to the H2–H5 bundle, where the bound actin-binding domain (ABD) may more likely adopt conformations similar to the five-helix weak state (dotted lines). Also see Xu et al., 2020.