Mechanism of the cadherin–catenin F-actin catch bond interaction

  1. Amy Wang
  2. Alexander R Dunn
  3. William I Weis

  • Curated by eLife

    eLife logo

    Evaluation Summary:

    This work investigated the mechanism of the cadherin-catenin F-actin catch bond interaction, a fundamental cell-cell adhesive structure that can be both dynamic and force-activated. Force measurements with purified protein components demonstrate that the catch bond results from a force-dependent switch of the actin-binding domain of αE-catenin between a five-helix bundle and a four-helix bundle bound on F-actin. The findings are interesting and well supported by experimental data, and will be interesting to the broader field of cytoskeleton function and functional structural biology.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Mechanotransduction at cell–cell adhesions is crucial for the structural integrity, organization, and morphogenesis of epithelia. At cell–cell junctions, ternary E-cadherin/β-catenin/αE-catenin complexes sense and transmit mechanical load by binding to F-actin. The interaction with F-actin, described as a two-state catch bond, is weak in solution but is strengthened by applied force due to force-dependent transitions between weak and strong actin-binding states. Here, we provide direct evidence from optical trapping experiments that the catch bond property principally resides in the αE-catenin actin-binding domain (ABD). Consistent with our previously proposed model, the deletion of the first helix of the five-helix ABD bundle enables stable interactions with F-actin under minimal load that are well described by a single-state slip bond, even when αE-catenin is complexed with β-catenin and E-cadherin. Our data argue for a conserved catch bond mechanism for adhesion proteins with structurally similar ABDs. We also demonstrate that a stably bound ABD strengthens load-dependent binding interactions between a neighboring complex and F-actin, but the presence of the other αE-catenin domains weakens this effect. These results provide mechanistic insight to the cooperative binding of the cadherin–catenin complex to F-actin, which regulate dynamic cytoskeletal linkages in epithelial tissues.

Article activity feed

  1. Version published to 10.7554/elife.80130 on eLife
  2. eLife

    Evaluation Summary:

    This work investigated the mechanism of the cadherin-catenin F-actin catch bond interaction, a fundamental cell-cell adhesive structure that can be both dynamic and force-activated. Force measurements with purified protein components demonstrate that the catch bond results from a force-dependent switch of the actin-binding domain of αE-catenin between a five-helix bundle and a four-helix bundle bound on F-actin. The findings are interesting and well supported by experimental data, and will be interesting to the broader field of cytoskeleton function and functional structural biology.

    (This preprint has been reviewed by eLife. We include the public reviews from the reviewers here; the authors also receive private feedback with suggested changes to the manuscript. Reviewer #1 agreed to share their name with the authors.)

  3. eLife

    Reviewer #1 (Public Review):

    In this work the authors investigated the mechanism of the cadherin-catenin F-actin catch bond interaction. They demonstrate that the catch bond is from a force-dependent switch of the ABD of αE-catenin between a five-helix bundle and a four-helix bundle bound on F-actin. In addition, they report cooperative binding of cadherin-catenin complexes on F-actin via interactions of neighboring ABDs. Overall, the findings are very interesting, the experiments are well executed, and the conclusions are backed up with experimental data.

    A notable strength of the work is the combination of single-molecule assay, analysis based on structural features, and kinetic modelling. Their proposal that the mechanism may be conserved across actin binding domains of several other actin binding proteins is convincing.

    The explanation of the observed catch bond is based on increased stability of the bound four-helix ABD compared to five-helix ABD. While this mechanism can explain the observed catch bond, alternative or additional factors need to be discussed. When the bound ABD switches from a five-helix bundle to a four-helix bundle, the stretching geometry is significantly altered. It is well known that different stretching geometries applied to the same molecule could lead to very different mechanical stability.

    It will also be helpful to add a brief discussion on how the results provide an understanding of the mechanical stability of the cadherin/beta-catenin/alpha-catenin/F-actin force-transmission linkage. At ~ 4 pN where the lifetime is the longest, the lifetime of the F-actin bound catenin complex is shorter than 10 s. Can the time scale enable robust mechanotransduction function? The time scale should sufficient to expose the vinculin binding site. After vinculin binds and engages with the same F-actin, will it stabilize (i.e., increase the lifetime) the linkage?

  4. eLife

    Reviewer #2 (Public Review):

    This group previously discovered that the cadherin/b-cat/a-cat "ternary " complex binds F-actin more strongly when the system is subjected to mechanical force using a single molecule/optical-trap-based approach (Buckley, 2014). A subsequent study rationalized the nature of a-cat's catch-bond behavior by modeling the a-cat actin-binding domain (ABD) alone (using CryoEM and a-cat F-actin sedimentation studies) to show that the first 1.5 helices of a-cat's 5-helical ABD (H0-H1) restricts high affinity F-actin binding (Xu et al., 2020). Whether this model held for a full length a-cat within a ternary complex, and whether removal of H0-H1 abrogated force-dependent F-actin binding using this group's established optical-trap/force measurements remained untested.

    MAJOR FINDING:

    1. The current study shows that removal of the N-terminal-most portion of a-cat's ABD (e.g., H0/H1; aa666-696) produces a cadherin-catenin complex that deviates from the normal, two-state binding seen for WT a-cat. Instead, the ternary ΔH0-H1 a-cat shows highest binding at the lowest forces and slips with increasing force. Using similar low-force binding conditions, the ternary ΔH0-H1 a-cat showed lifetime binding 39x longer than WT. Ternary ΔH0-H1 and WT show no difference in binding to F-actin subjected to forces above 6pN. Thus, the Ternary ΔH0-H1 behaves as single-state slip bond (compared to WT showing a two-state binding). This finding affirms a model put forth by this same group (Xu et al., 2020), suggesting that undocking/removal of H1 converts a 5-helical bundle ABD associated with weak-actin binding to a 4-helical ABD with stronger binding/longer lifetime. STRENGTH: This finding is important because a previous study assigned force-sensitive actin-binding to the H0-region of the 5-Helical ABD, but it is clear that the H1 region is most critical to this force-gating mechanism. This finding better rationalizes how other ABD that lack H0 show force-dependent binding to F-actin (e.g., vinculin, talin).

    OTHER NOTABLE FINDINGS:

    2. Evidence that aCat ABD alone binds F-actin 4x-longer than a full length aCat within the ternary complex suggests presence of cadherin/b-cat or aCat N and M-domains can negatively impact ABD. This fits with other studies from this group (Drees et al., 2005; Terekhova et al., 2019).
    3. Evidence addition of purified ABD to Ternary complex/optical trap assay increased binding life-time by 4x is interesting and consistent with previous work. Given so much attention given to the CTE in the final model (Fig. 6), I am (a bit) surprised the investigators did not use their CTE-deletion mutants from Xu et al., 2020, to validate the molecular nature of this cooperativity (or contribution to the force-dependent binding). But answering this question is not necessary to interpret other aspects of the manuscript.
    4. Evidence that H0/H1 imposes directionality to a-cat/F-actin catch-bond mechanism is intriguing: WT a-cat favors binding to actin filaments pulled towards the minus (non-growing) rather than + (growing) end, whereas a-cat lacking H0/H1 fails to show this bias. However, given most actin structures that interface with junction structures are of mixed polarity, I struggle to understand the broader meaning of this finding. While not the job of this team to give up their working hypotheses and means of testing the broader significance of this finding, some clarity here would help.

  5. Version published to 10.1101/2022.05.06.490842v1 on bioRxiv