The tale of two talins - two isoforms to fine-tune integrin signalling

Abstract

Talins are cytoplasmic adapter proteins essential for integrin-mediated cell adhesion to the extracellular matrix. Talins control the activation state of integrins, link integrins to cytoskeletal actin, recruit numerous signalling molecules that mediate integrin signalling and coordinate recruitment of microtubules to adhesion sites via interaction with KANK (kidney ankyrin repeat-containing) proteins. Vertebrates have two talin genes, TLN1 and TLN2. Although talin1 and talin2 share 76% protein sequence identity (88% similarity), they are not functionally redundant, and the differences between the two isoforms are not fully understood. In this Review, we focus on the similarities and differences between the two talins in terms of structure, biochemistry and function, which hint at subtle differences in fine-tuning adhesion signalling.

Keywords: integrin; mechanobiology; talin.

Figure 1

Talin at the core of the adhesion. A cartoon of the core of integrin adhesions, highlighting talins central role. Talin coordinates both the actin cytoskeleton, and through the interaction with KANK proteins, the microtubule cytoskeleton at adhesion sites. Once the adhesion core is assembled, talin serves as a scaffold to recruit many other proteins in order to form all the many different types of adhesive structures (focal adhesions, podosomes, invadopodia, etc.).

Figure 2

Structure and domain map of the two talin isoforms. (A) Structural model of talin showing the domain arrangement of talin. Vinculin‐binding sites are shown in red. The N‐terminal talin head comprising F0–F3 and the talin rod domains R1–R13 are shown. (B) Schematic representation of the talin domain structures coloured by sequence identity between the two isoforms. The domain boundaries are given for mouse talin1 (UniProt: http://www.uniprot.org/uniprot/P26039) and talin2 (UniProt: http://www.uniprot.org/uniprot/B2RY15). Provided that these boundaries are used, it is possible to make any talin fragment or delete any talin domain while maintaining the structural integrity of the protein. (C) The locations of many of the talin ligand‐binding sites are shown, as are the calpain cleavage sites.

Figure 3

Layers and layers of autoinhibition. A striking feature of talins is their remarkable conformational plasticity that enables different ligands to engage the same platform in different conditions; part of this flexibility emerges from autoinhibition. (Left) In the closed autoinhibited form, all of the domains are folded, and many of the ligand‐binding sites for actin, integrin and vinculin are thought to be cryptic. Some binding sites may face outwards and remain accessible; for instance, RIAM is able to bind to the inactive conformation 127. In the extended conformation in the absence of force, all the domains are still folded, and additional binding sites are exposed (IBS1, IBS2, ABS3, plus the sites for those ligands that require folded‐rod domains) (Right). The exposure of IBS1 and ABS3 facilitates adhesion formation, and by activating integrins and crosslinking them to the actin cytoskeleton, a nascent adhesion can form. As force is exerted on talin, another layer of autoinhibition is uncovered (Bottom). As talin domains unfold, starting with R3, the initial mechanosensor in talin 21, 23, 24, vinculin‐binding sites are exposed and talin:vinculin interactions can now occur. R3 unfolding also reveals the high affinity actin‐binding site in talin, ABS2 that can then activate tension‐bearing actin connections 84, 85. As domains unfold, the binding sites for ligands that engage the folded rod domains are destroyed, as is the case for RIAM binding to R3. A remarkable feature of talins conformational plasticity is that, in the absence of other factors, talin can readily refold to its default low‐force state.