The Talin FERM Domain Is Not So FERM

CORE Metadata, citation and similar papers at core.ac.uk

Provided by Elsevier - Publisher Connector Structure Previews

a thermosome in the open state is there- complement to the recent reports on Huo, Y., Hu, Z., Zhang, K., Wang, L., Zhai, Y., Zhou, fore of interest and will help to complete group II chaperonins by Zhang et al. Q., Lander, G., Zhu, J., He, Y., Pang, X., et al. (2010). Structure 18, this issue, 1270–1279. our understanding of the mechanism of (2010) and Pereira et al. (2010), and builds group II chaperonins. From their structural upon their impressive work by providing Pereira, J.H., Ralston, C.Y., Douglas, N.R., Meyer, D., Knee, K.M., Goulet, D.R., King, J.A., Frydman, analysis of the open state, they show a substantially higher resolution structure 285 J., and Adams, P.D. (2010). J. Biol. Chem , a rotation of 30 of the apical and lid of the group II chaperonin in the open 27958–27966. domains relative to the closed state, state. providing the clearest picture yet of the Shomura, Y., Yoshida, T., Iizuka, R., Maruyama, T., Yohda, M., and Miki, K. (2004). J. Mol. Biol. 335, domain movements resulting from ATP 1265–1278. hydrolysis. The authors additionally report REFERENCES electron microscopy reconstructions for Zhang, J., Baker, M.L., Schroder, G.F., Douglas, Ditzel, L., Lowe, J., Stock, D., Stetter, K.-O., Huber, N.R., Reissmann, S., Jakana, J., Dougherty, M., both the open and closed states of the H., Huber, R., and Steinbacher, S. (1998). Cell 93, Fu, C.J., Levitt, M., Ludtke, S.J., et al. (2010). thermosome. This work provides a nice 125–138. Nature 463, 379–384.

Iain D. Campbell1,* 1Department of Biochemistry, University of Oxford, South Parks Road, Oxford, OX1 3QU, UK *Correspondence: [email protected] DOI 10.1016/j.str.2010.09.002

The structure of the head domain of talin, an intracellular activator of integrin membrane adhesion receptors, has been solved by Elliott et al. (2010). A FERM domain can be identiﬁed in the head from sequence compar- isons but, rather than having a compact structure of three subdomains, it has linear arrangement of four subdomains.

FERM domains are found in numerous talin is reported (Elliott et al., 2010). This adhesion receptors (Campbell and Gins- proteins located at the cytoplasmic face has an unexpected structure with a linear, berg, 2004). Talin has an N-terminal of the plasma membrane (Fehon et al., rather than a clover-leaf, arrangement of head region and an elongated 220 kDa 2010). The FERM name derives from its subdomains. helical rod that combine to link the cyto- presence in four proteins: band four-point Talin is an intracellular protein that is plasmic tail of the b-integrin subunit one, ezrin, radixin, and moesin. Other a key player in the activation of integrins, with the actin cytoskeleton (Critchley, important FERM-containing proteins in- large heterodimeric membrane-spanning 2009). The head contains a FERM domain clude focal adhesion kinase with clear sequence similarities (FAK) and Janus kinase to other FERM proteins, (JAK). FERM domains have although the F1 subdomain around 300 amino acids has a 30-residue insertion and with three subdomains, some of the linker regions usually called F1, F2, and are different. The talin FERM F3. Several structures of domain is also preceded by FERM domains have been an ‘‘F0’’ subdomain, recently solved; there is some varia- shown to have an ubiquitin- tion, especially in linker like fold, similar to F1 (Goult regions and loop insertions, et al., 2010). The 30-residue but all previous structures insertion in F1 is largely have had a relatively unstructured, but it has helical compact clover-leaf struc- Figure 1. Subdomain Arrangements in FERM Proteins propensity and can be removed ture with intimate contacts An illustration of the clover-leaf arrangement of subdomains observed in without perturbing the core between all three subdo- previous FERM structures (left) and the novel arrangement found in the structure of F1 (Goult et al., mains (Figure 1). In this FERM domain from the talin head (right). Positive patches on the F1 and F2 2010). Many unsuccessful subdomains can interact with negatively charged membranes. F3 binds the issue, the crystal structure cytoplasmic tail of the integrin b-subunit (Anthis et al., 2009) and F0-F3 then attempts have been made to of the N-terminal head of act in synergy to activate integrins. obtain structures of the intact

talin head region but Elliott et al. (2010) A number of other interesting features REFERENCES achieved success by removing the flex- emerge from the new talin head structure. ible F1 insertion. Their data from the FERM domain activity is often regulated Anthis, N.J., Wegener, K.L., Ye, F., Goult, B.T., Lowe, E.D., Vakonakis, I., Bate, N., Critchley, crystal structure were supplemented by by various inter- and intramolecular inter- D.R., Ginsberg, M.H., and Campbell, I.D. (2009). X-ray scattering and NMR studies on the actions (Fehon et al., 2010). Talin is also EMBO J. 28, 3623–3632. intact head. The results suggest that the autoinhibited by intermolecular associa- Campbell, I.D., and Ginsberg, M.H. (2004). Trends F0-F1 and F2-F3 subdomain pairs form tion between the F3 subdomain and Biochem. Sci. 29, 429–435. relatively rigid structures, while the F1- a helical bundle in the rod region (Goksoy Critchley, D.R. (2009). Annu. Rev. Biophys. 38, F2 linker is flexible. et al., 2008; Goult et al., 2009a). The new 235–254. The F3 subdomain, which can be structure shows that this interaction Elliott, P.R., Goult, B.T., Kopp, P.M., Bate, N., classified as a PTB (phospho-tyrosine would not only mask the integrin binding Grossmann, J.G., Roberts, G.C.K., Critchley, binding) domain, is often the key binding site in F3, but also prevent the F2 and F3 D.R., and Barsukov, I.L. (2010). Structure 18, this site for FERM domain interactions with domains interacting with the membrane; issue, 1289–1299. other proteins; talin F3 is no exception, the structure suggests, however, that the Fehon, R.G., McClatchey, A.I., and Bretscher, A. as shown by its ability to bind to b-integrin F1 and F0 domains would be still be able (2010). Nat. Rev. Mol. Cell Biol. 11, 276–287. tails. A basic patch on the F2 subdomain to locate the autoinhibited talin within the Goksoy, E., Ma, Y.-Q., Wang, X., Kong, X., Perera, docks on the membrane, helping to membrane. D., Plow, E.F., and Qin, J. (2008). Mol. Cell 31, 124–133. orient the membrane-spanning helix of The kindlin family of proteins, which the b-integrin subunit and to promote assist talin in activating integrins (Moser Goult, B.T., Bate, N., Anthis, N.J., Wegener, K.L., a b Gingras, A.R., Patel, B., Barsukov, I.L., Campbell, separation of the and integrin subunits et al., 2009), have FERM domains with I.D., Roberts, G.C.K., and Critchley, D.R. (2009a). (Anthis et al., 2009)(Figure 1). Elliott et al. talin-like features, such as an N-terminal J. Biol. Chem. 284, 15097–15106. (2010) show here that the positive patch F0 domain, a large F1 loop (Goult et al., Goult, B.T., Bouaouina, M., Harburger, D.S., Bate, on F2 and the positively charged F1 inser- 2009b), and an F1-F2 linker that is N., Patel, B., Anthis, N.J., Campbell, I.D., Calder- tion loop facilitate cell spreading in cells predicted to be flexible. This sug- wood, D.A., Barsukov, I.L., Roberts, G.C., and 394 transfected with modified talin heads. gests that the subdomain arrangement Critchley, D.R. (2009b). J. Mol. Biol. , 944–956. The extended conformation of the talin seen in the talin FERM domain may Goult, B.T., Bouaouina, M., Elliott, P.R., Bate, N., head thus seems to have evolved to not be an unusual outlier; it may be Patel, B., Gingras, A.R., Grossmann, J.G., Roberts, G.C.K., Calderwood, D.A., Critchley, D.R., and bind both to integrins, via F3, and to nega- the first of a number of proteins with Barsukov, I.L. (2010). EMBO J. 29, 1069–1080. tively charged microdomains in the lipid this noncanonical arrangement of sub- Moser, M., Legate, K.R., Zent, R., and Fa¨ ssler, R. bilayer (Figure 1). domains. (2009). Science 324, 895–899.

DEEP Insights through the DEP Domain

Wenqing Xu1,* and Xi He2,* 1Department of Biological Structure, University of Washington, Seattle, WA 98195, USA 2F.M. Kirby Neurobiology Center, Children’s Hospital Boston, Harvard Medical School, Boston, MA 02115, USA *Correspondence: [email protected] (W.X.), [email protected] (X.H.) DOI 10.1016/j.str.2010.09.007

In this issue, Yu et al. (2010) provide a crystal structure for the bipartite interface between the m2 subunit of the adaptor protein AP-2 complex and Dishevelled, a key component for Wnt signaling.

Secreted Wnt proteins regulate embryo- is required downstream of Wnt/Fz for both (Dishevelled/Axin), PDZ (PSD-95, DLG, genesis and homeostasis by activating these pathways (Gao and Chen, 2010; ZO1), and DEP (Dishevelled, EGL-10, multiple intracellular signaling pathways, MacDonald et al., 2009). But how Dvl acti- Pleckstrin) (Gao and Chen, 2010). Sim- including the canonical b-catenin and the vates these distinct downstream path- plistically speaking, the N-terminal DIX noncanonical planar cell polarity (PCP) ways remains enigmatic. Some recent domain of Dvl functions mainly in canon- pathways. Frizzled (Fz) proteins, the main insights, including a report in this issue ical signaling, and the central PDZ domain type of Wnt receptors, together with other of Structure (Yu et al., 2010), help to is required in both pathways via interac- coreceptors mediate activation of the shed light on this long-standing question. tion with the cytoplasmic tail of Fz, while Wnt/b-catenin and/or PCP pathways. Dvl is a scaffold protein containing the more carboxyl DEP domain is critical The cytoplasmic Dishevelled (Dvl) protein three highly conserved domains: DIX for PCP signaling via mediation of the