Force transmission by talin and kindlin at integrin tail domains
Fusion constructs between the integrin-binding domain 3 of talin and the beta integrin tail were designed and DNA handles added to the ends of this construct. Single molecule optical tweezers experiments showed that the talin-integrin bond is mechanically much weaker than expected in the current cell biological models. However, adding kindlin protects the talin-integrin bond and renders it force insensitive. This discovery points towards a mechanism where a tripartite complex between integrin, talin, and kindlin enhances the mechanical stability of the connection between cytoskeleton and extracellular matrix in the early events of adhesion formation. Further work with a number of variants of talin and integrin showed that this effect depends on the integrin isoform and is present in beta 1d but not in beta1a integrin subunit. This work has been published in PNAS (Bodescu et al., 2023).
Role of integrin-actomyosin linkage for kidney development
By combining NMR, biochemistry and cell biology, we discovered that talin and kindlin binding to the β1-integrin tail (β1-CT) can induce a conformational change in the β1-CT that increases talin and decreases kindlin affinity. We also discovered that this asymmetric affinity regulation is accompanied by a direct interaction between talin and kindlin, which can promote the simultaneous binding of talin and kindlin to β1-CTs. Disrupting either the allosteric communication between the binding sites of talin and kindlin or their direct interaction in cells severely compromised integrin functions. A revised manuscript describing the study has been resubmitted for publication.
Based on this finding, we used the results from the NMR line broadening experiment of our study to perform an alanine scan of the β1-CT in regions that were extraordinarily perturbed in ternary complex titrations to identify residues that might play a role in the allosteric communication between talin and kindlin. First, we measured the affinity of kindlin2 towards these mutants. We did not identify significant changes in affinity for kindlin suggesting that the allosteric coupling is activated upon simultaneous binding of talin and kindlin. Second, we measured the talin affinity for the β1-CT mutants in the absence and presence of kindlin2 and found that certain substitutions between the talin and kindlin NPxY motifs reduce talin affinity in the presence of kindlin. The most interesting pair of mutations that we discovered in this mutational analysis was β1-CT-S785A and -S785D. Whereas the apparent talin affinity for β1-CT-S785A was unaffected by the presence of kindlin2, the apparent THD1 affinity for β1-CT-S785D significantly decreased by the presence of kindlin2.
For a long time we assumed that β1-CT-S785 is bound and phosphorylated by an unknown kinase. However, crosslinking proteomics and an siRNA screen of all known kinases did not produce hits indicating that not phosphorylation but the conformational change of the tail by the β1-CT-S785D affects talin binding. To test the in vivo consequences of the talin affinity change for β1-CT-S785D, we generated a mouse model carrying the β1-S785D substitution in the germline. Whereas mice carrying the β1-S785A substitution are perfectly normal, mice carrying β1-S785D mice lack kidneys. Our preliminary data indicate that the ureteric bud outgrowth is severely compromised in β1-S785D mice. Importantly, we could confirm with a series of in vivo and ex vivo experiments that the kidney defect is caused by reduced talin affinity and not by integrin tail modification such as serine-785 phosphorylation. The studies are completed and we are in the process of finalizing a manuscript.
In parallel to the kidney model, other organoid systems were tested. Single cell derived organoid systems of human mammary gland and murine pancreatic tumor were identified to be the most effective way to move forward to establish the role of cell adhesion on the self organization of epithelial cell layers.
