Available online at www.sciencedirect.com
ScienceDirect Signalling complexes at the cell-matrix interface Erhard Hohenester The extracellular matrix critically controls cell behaviour. Many cell-matrix interactions are mediated by transmembrane receptors of the integrin family. In the last two years, the structural changes resulting from ligand binding to integrins a5b1, avb3 and aIIbb3 have been mapped in unprecedented detail. The structure of integrin aXb2 has revealed how ligand binding to the a I domain is transmitted to the rest of the ectodomain. The structural characterisation of the cytosolic regulator talin has been continued, revealing how the integrin binding site is blocked in auto-inhibited talin. Finally, structures of the discoidin domain receptors DDR1 and DDR2 have begun to reveal how these atypical receptor tyrosine kinases become activated by the major matrix component collagen. Addresses Department of Life Sciences, Imperial College London, Sir Ernst Chain Building, London SW7 2AZ, UK Corresponding author: Hohenester, Erhard ([email protected])
Current Opinion in Structural Biology 2014, 29:10–16 This review comes from a themed issue on Multi-protein assemblies in signalling Edited by Deborah Fass and Katrin Rittinger
http://dx.doi.org/10.1016/j.sbi.2014.08.009 0959-440X/# 2014 Elsevier Ltd. All right reserved.
Introduction Most cells in higher animals are organised into tissues, which consist of differentiated cells in contact with each other and/or with extracellular matrix (ECM). The ECM is composed of large glycoproteins, which frequently are assembled into supramolecular structures [1,2]. Sheetlike matrices termed basement membranes cover the basal side of all epithelia and surround muscle, fat and peripheral nerve cells. Fibrillar matrices impart tensile strength (collagen fibres) and elasticity (elastic fibres) to many tissues. In addition to providing structural support for cells, the ECM controls many fundamental cellular processes, such as cell survival, polarisation, differentiation and migration. Some of these effects are achieved by modulating the diffusion and availability of growth factors, which signal through enzyme-linked transmembrane receptors. However, many signals derive from the composition and biomechanical properties of the ECM per se. Current Opinion in Structural Biology 2014, 29:10–16
Matrix stiffness is now recognised as a powerful regulator of cell differentiation, for instance, but how stiffness regulates gene expression remains incompletely understood [3]. A major hub of cell-matrix signalling are focal contacts and adhesions, in which transmembrane receptors of the integrin family link ECM molecules to a complex and dynamic network of intracellular adaptors and the cytoskeleton [4,5].
Integrin structure and shape-shifting The 24 mammalian integrins are assembled from 18 a subunits and eight b subunits, and include receptors for the major ECM proteins fibronectin (e.g. a5b1, avb3), laminins (e.g. a3b1, a6b1, a7b1) and collagens (e.g. a1b1, a2b1) [6]. Integrins have a large extracellular region consisting of a ligand binding head and two legs, which are connected to single transmembrane helices and short cytoplasmic domains (Figure 1a). The globular head is constructed from both subunits: the a subunit contributes a b-propeller domain, which in some integrins contains an inserted (a I) domain; the b subunit contributes the b I domain (so called because of its homology to the a I domain). When present, the a I domain contains the ligand binding site. In all other integrins, the ligand is believed to bind at the a/b interface with major determinants residing in the b I domain. Ligand binding involves metal ion-dependent adhesion sites (MIDAS) in both a and b I domains and triggers similar conformational transitions within the I domains (most importantly, a piston-like movement of the C-terminal helix a7) [7]. The first crystal structure of a complete integrin ectodomain (avb3) revealed a severely bent conformation that is now believed to represent a state with low affinity for ligand (Figure 1a) [8,9]. A dramatically different conformation with straight and separated legs was observed by electron microscopy and interpreted as the high-affinity state [10]. A pertinent feature of the high-affinity conformation, namely the swing-out of the b hybrid domain (Figure 1a), was observed in a crystal structure of the ligand-binding headpiece of platelet integrin aIIbb3 [11]. These and many other studies strongly support an allosteric mechanism of integrin activation [6,7], but the delicate energetic balance of integrin shape-shifting continues to throw up surprises (see below). Importantly, a recent single particle reconstruction of full-length aIIbb3 integrin embedded in nanodiscs suggest that the physiological resting state may differ substantially from the bent conformation observed in several crystal structures [12]. Given the demonstrated importance of the transmembrane region in integrin activation [13], further structural www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 11
studies of integrins in their native environment are clearly warranted. Several recent studies have shown that bent integrins in the crystalline state can bind ligand and display structural features typical of activated integrins. The structure of the a5b1 integrin headpiece (the first structure of a member of the large and ancestral b1 integrin subfamily) was determined in the low-affinity (closed) conformation in complex with an inhibitory allosteric antibody [14]. Similarly to the bent avb3 and aIIbb3 integrins, the closed a5b1 integrin headpiece nevertheless bound a fibronectin-mimetic Arg-Gly-Asp (RGD) peptide. This information allowed the in silico docking of the integrinbinding FN9-FN10 region of fibronectin, which helped identify the residues in the integrin a5 subunit that interact with the so-called synergy site in FN9 [14]. In another study, a crystal structure of avb3 integrin with fibronectin FN10 was obtained by soaking crystals of bent avb3 integrin with FN10 protein. FN10 binding was associated with structural changes in the b I domain, which were not observed with a purely antagonistic FN10 mutant [15]. Similarly, soaking of RGD peptides into crystals of the closed aIIbb3 headpiece allowed various intermediate states on the pathway to the highaffinity conformation to be captured, including a structure with a partially swung out b3 subunit hybrid domain [16]. The collagen-binding integrins a1b1, a2b1, a10b1 and a11b1 all contain an a I domain [17] that responds to collagen binding with a downward movement of the Cterminal a7 helix [18]. How this change is transmitted to the remainder of the integrin ectodomain, and ultimately the cytosolic domains, has been unclear. A structure of the leucocyte integrin aXb2 has provided the first detailed view of the activation mechanism of a I domain-containing integrins [19]. Even though the overall conformation of aXb2 integrin was bent, the a I domain was captured in the high-affinity conformation with the a7 helix in the downward position (Figure 1b). The segment following the a7 helix contains a conserved glutamic acid (Glu318 in aXb2) that had previously been hypothesised to provide an internal ligand for the b I domain [20]. In the new crystal structure, Glu318 indeed coordinates the MIDAS magnesium ion in the b I domain, thus priming the transition to the active conformation of the b I domain and the eventual swing-out of the hybrid domain [19].
Cytoplasmic interactions with integrins The conformational changes in the integrin heterodimer are used to transmit signals across the plasma membrane in both directions [6,7]. In outside-in signalling, ligand binding to the ectodomain leads to a separation of the two legs, which allows the b subunit cytoplasmic domain to bind intracellular proteins such as talin and kindlins. In inside-out signalling, intracellular activation of talin www.sciencedirect.com
(discussed below) leads to its recruitment to the plasma membrane, where it can bind to the b subunit cytoplasmic domain, thereby triggering the transition of the integrin heterodimer to a state with high affinity for extracellular ligand. How talin binds to the integrin b subunit and activates integrins is now well understood, but a molecular explanation for kindlins’ essential roles in integrin signalling is currently lacking [21–25]. A recent study has shown that kindlins are required for the clustering of talin-activated integrin [26]. Talin and kindlins have in common an N-terminal atypical FERM (4.1 protein, ezrin, radixin, moesin) domain that mediates integrin binding [27–30]. In talin (but not in kindlins) this so-called head domain is followed by a long rod consisting of 13 a-helix bundles and terminating in a dimerisation domain (Figure 2a) [22]. The structural characterisation of the rod was recently completed in a tour de force, in which NMR spectroscopy was used to define folded units before detailed structure determination [31]. The a-helix bundles of talin contain multiple binding sites for the actin-binding protein vinculin. Vinculin binding is mediated by the hydrophobic surfaces of amphipathic helices in talin, which become exposed in a process of a-helix bundle conversion [32,33]. A singlemolecule study has shown that mechanical stretching of the talin rod activates vinculin binding, providing a mechanism for force-sensing at focal adhesions [34]. Vinculin also needs to be restructured in order to bind talin, but there is disagreement about the precise mechanism of vinculin activation [32,35,36]. In contrast to the isolated talin rod, in which the flexibly linked a-helix bundles form an extended array, fulllength (dimeric) talin assumes a compact conformation that represents an autoinhibited form with regard to integrin binding. A recent single particle reconstruction of full-length talin places the two heads of the dimer inside a doughnut-shaped structure formed by the rods [37]. A key feature of this assembly is the interaction of the F3 subdomain of the talin head with the ninth a-helix bundle, R9. The crystal structure of a talin F2F3-R9 complex showed that the rod segment not only occludes the integrin binding site in F3, but also antagonises a positively charged surface in F2 that mediates the interaction of the talin head with the plasma membrane (Figure 2b,c) [38]. While the reductionist approach of traditional structural investigations has been very successful in defining the key interactions in integrin-mediated cell-matrix contacts, a complete picture will only emerge from structural studies of focal adhesions in situ [4]. Super-resolution fluorescence microscopy has been used to determine the vertical organisation of focal adhesions formed by cells plated on fibronectin-coated glass [39]. The layer closest to the plasma membrane contains the integrin Current Opinion in Structural Biology 2014, 29:10–16
12 Multi-protein assemblies in signalling
Figure 1
Ligand
(a) Ligand
α I domain Conserved Glu
Propeller domain
β I domain Hybrid domain
Thigh domain
α
α
β
Closed headpiece, bent
PSI domain
α
β
α
β
Open headpiece, extended
Closed headpiece, extended
β
α I domain-containing integrin
MIDAS
(b)
αI
βI
Propeller
Thigh
Hybrid
PSI
α
β Current Opinion in Structural Biology
Integrin structure and conformational changes. (a) Schematic drawings of integrins in different functional states (adapted from [7]). The coloured domains represent actual crystal structures (except for the situation on the right), whereas the locations of the grey domains are inferred from electron microscopic reconstructions. The plasma membrane is represented by a grey bar. During headpiece opening, a downward movement of the Cterminal helix of the b I domain (magenta) is coupled to a swing-out of the hybrid and plexin-semaphorin-integrin (PSI) domains. In integrins with an a I domain, a similar movement of the C-terminal helix in the a I domain (red) results in internal ligation of the b I domain metal ion-dependent adhesion site (MIDAS), which in turn leads to hybrid domain swing-out. (b) Crystal structure of integrin aXb2 in a metastable conformation [19]. Only the headpiece portion is shown for clarity. Calcium and magnesium ions are shown as black and yellow spheres, respectively. The a I domain is in the active conformation, with the C-terminal helix in the down position and the following segment (red) ligating the b I domain MIDAS. The C-terminal helix of the b I domain (magenta) is trapped in an intermediate position and the headpiece is closed. The positions of the hybrid and PSI domains in the open conformation, as observed in an integrin aIIbb3 structure [11], are indicated in grey.
Current Opinion in Structural Biology 2014, 29:10–16
www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 13
Figure 2
(a)
Head
Rod R8
R4 R2 F0 F1 F2 F3
R5
R6
R7
R9
R10
R11
R12
R13
DD
R1 R3
(b)
(c) Negative patch + + + +
Integrin β1 cytoplasmic region
R9 F2 F3
Current Opinion in Structural Biology
Talin structure and autoinhibition. (a) Domain structure of talin. The talin head consists of a FERM (4.1 protein, ezrin, radixin, moesin) domain with linearly arranged subdomains F0–F4. The talin rod consists of 13 a-helix bundles [31]. R1-R4 and R8 are 4-helix bundles, which have their N-termini and C-termini close together, whereas the remaining bundles have five helices, allowing a linear head-to-tail arrangement. Full-length talin is dimerised via the C-terminal dimerisation domain (DD) and adopts a compact autoinhibited conformation [37] that is converted to a more open conformation by a variety of activators. The talin rod contains multiple binding sites for the actin-binding protein vinculin (marked by yellow stars) that become exposed by mechanical tension. (b) Crystal structure of the talin-2 F2F3 domains bound to the cytoplasmic region of integrin b1 [27]. The plasma membrane is represented by a grey bar. A positive patch on the F2 domain interacts with the phospholipid headgroups. (c) Crystal structure of the talin-1 F2F3 domains bound to the rod segment R9, representing the autoinhibited state of talin [38]. R9 blocks the integrin binding site and contains a negative patch that disfavours membrane binding.
cytoplasmic domains, the talin head domain, paxillin and focal adhesion kinase. This is followed by a ‘force transduction layer’ composed of the talin rod domain and vinculin. Actin was detected only at greater than 40 nm distance from the plasma membrane [39].
and sustained. The DDRs do not mediate firm cell adhesion themselves, but they influence adhesion indirectly by regulating the activation state of collagen-binding integrins [43]. However, in common with typical RTKs, the main functions of DDRs appear to be related to signalling and the control of cell proliferation [17,42].
Discoidin domain receptors Integrins are not the only receptors for ECM molecules. A particularly intriguing class of non-integrin ECM receptors are the two discoidin domain receptors, DDR1 and DDR2, which are receptor tyrosine kinases (RTKs) that are activated by collagen [17,40–42]. Compared with typical RTKs that are activated by small diffusible growth factors, DDR activation by collagens is slow www.sciencedirect.com
DDRs consist of an N-terminal discoidin (DS) domain that contains the collagen binding site, followed by a DSlike domain [44], a long extracellular juxtamembrane (JM) region (30–50 residues), a single transmembrane helix, an even longer cytoplasmic JM region (up to 160 residues), and a canonical tyrosine kinase domain. Systematic investigations using collagen peptide libraries Current Opinion in Structural Biology 2014, 29:10–16
14 Multi-protein assemblies in signalling
identified a GVMGFO motif (O is hydroxyproline) as the major DDR binding site in fibrillar collagens; this site is distinct from all known integrin binding sites [45–47]. Triple-helical peptides containing the GVMGFO motif bind to an amphipathic pocket at the top of the DS domain of DDRs (Figure 3) [48,49]. How collagen binding leads to DDR activation (i.e. autophosphorylation of the cytoplasmic regions) is still unclear. Typical RTKs respond to ligand binding by dimerisation or by structural changes within a constitutive dimer [50,51]. In either case, the JM regions frequently play an important role in activation by restraining the inactive RTK and by transmitting structural changes across the plasma membrane [52,53]. DDRs are constitutive noncovalent dimers [54]. A recent study showed that the extracellular JM region of DDR1 is extremely tolerant to mutation: DDR1 activation was unaffected by disulphide cross-links, deletions or insertions of flexible segments [55]. These findings argue against a mechanism of conformational coupling across the plasma membrane and are Figure 3
more consistent with signalling by DDR clusters, but information about DDR stoichiometry at the cell surface is currently lacking.
Concluding remarks Integrin-mediated adhesion and signalling is now quite well understood at the level of single integrin heterodimers. The challenge for the future is to integrate these insights with structural and functional studies using cultured cells and observations from animal experiments. One potential concern is that the prominent focal adhesions observed in cells plated on stiff two-dimensional substrates may not always be representative of the situation in three-dimensional matrices, which are less amenable to detailed structural analyses. Another interesting question is how well the ECM substrates that are used in vitro actually correspond to the situation in vivo. This is particularly troubling in the case of collagen binding by integrins and DDRs. Almost all experiments have been performed with isolated triple helices, but many collagens form highly organised fibrils, the surface of which is decorated with a variety of accessory ECM proteins. It will, therefore, be important to investigate how integrins and DDRs interact with native fibrillar collagens.
Collagen
Conflict of interest None declared.
Acknowledgements This review is dedicated to the memory of Iain D Campbell, who pioneered structural studies of integrin activation. I thank Birgit Leitinger and David Pulido for critical reading of the manuscript. EH is supported by the Wellcome Trust.
DS domain
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest
DS-like domain
JM region (50 aa)
Current Opinion in Structural Biology
Discoidin domain receptor structure. The figure shows a composite structure, consisting of the unliganded DDR1 ectodomain [44] and a collagen triple helix (orange) docked according to the crystal structure of the DDR2 discoidin (DS) domain bound to a collagen-like peptide [48]. A surface patch on the DS domain that is essential for DDR activation [44] is coloured magenta. Disulphide bridges and N-linked glycans are in yellow and light pink, respectively. A flexible 50-residue juxtamembrane (JM) region links the DS-like domain to the transmembrane region, which mediates the constitutive dimerisation of DDRs [54,55]. Current Opinion in Structural Biology 2014, 29:10–16
1.
Hynes RO: The extracellular matrix: not just pretty fibrils. Science 2009, 326:1216-1219.
2.
Bruckner P: Suprastructures of extracellular matrices: paradigms of functions controlled by aggregates rather than molecules. Cell Tissue Res 2010, 339:7-18.
3.
Discher DE, Janmey P, Wang YL: Tissue cells feel and respond to the stiffness of their substrate. Science 2005, 310:1139-1143.
4.
Hanein D, Horwitz AR: The structure of cell-matrix adhesions: the new frontier. Curr Opin Cell Biol 2012, 24:134-140.
5.
Wolfenson H, Lavelin I, Geiger B: Dynamic regulation of the structure and functions of integrin adhesions. Dev Cell 2013, 24:447-458.
6.
Hynes RO: Integrins: bidirectional, allosteric signaling machines. Cell 2002, 110:673-687.
7.
Luo BH, Carman CV, Springer TA: Structural basis of integrin regulation and signaling. Annu Rev Immunol 2007, 25:619-647.
8.
Xiong JP, Stehle T, Diefenbach B, Zhang R, Dunker R, Scott DL, Joachimiak A, Goodman SL, Arnaout MA: Crystal structure of the extracellular segment of integrin avb3. Science 2001, 294:339-345. www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 15
9.
Xiong JP, Stehle T, Zhang R, Joachimiak A, Frech M, Goodman SL, Arnaout MA: Crystal structure of the extracellular segment of integrin avb3 in complex with an Arg-Gly-Asp ligand. Science 2002, 296:151-155.
10. Takagi J, Petre BM, Walz T, Springer TA: Global conformational rearrangements in integrin extracellular domains in outside-in and inside-out signaling. Cell 2002, 110:511-599. 11. Xiao T, Takagi J, Coller BS, Wang JH, Springer TA: Structural basis for allostery in integrins and binding to fibrinogenmimetic therapeutics. Nature 2004, 432:59-67. 12. Choi WS, Rice WJ, Stokes DL, Coller BS: Three-dimensional reconstruction of intact human integrin aIIbb3: new implications for activation-dependent ligand binding. Blood 2013, 122:4165-4171. This paper reports the structure at 20 A˚ resolution of integrin aIIbb3 embedded in a nanodisc lipid bilayer. The ligand-binding headpiece in the low-affinity state is more accessible than predicted from crystal structures and the arrangement of the two legs is also different, leading the authors to propose modification of the switchblade model of integrin extension [7]. 13. Kim C, Schmidt T, Cho EG, Ye F, Ulmer TS, Ginsberg MH: Basic amino-acid side chains regulate transmembrane integrin signalling. Nature 2012, 481:209-213. This study shows that a ‘snorkelling’ lysine in the transmembrane region of the integrin b3 domain is required for correct membrane embedding. Mutation of this lysine causes integrin activation by changing the membrane-crossing angle. 14. Nagae M, Re S, Mihara E, Nogi T, Sugita Y, Takagi J: Crystal structure of a5b1 integrin ectodomain: atomic details of the fibronectin receptor. J Cell Biol 2012, 197:131-140. This paper reports the first structure of an integrin from the b1 subfamily. Modelling shows how integrin a5b1 binds fibronectin through a primary interaction with the RGD sequence in FN10 and a synergy site in FN9. 15. Van Agthoven JF, Xiong JP, Alonso JL, Rui X, Adair BD, Goodman SL, Arnaout MA: Structural basis for pure antagonism of integrin avb3 by a high-affinity form of fibronectin. Nat Struct Mol Biol 2014, 21:383-388. This paper reports that fibronectin FN10 binding to integrin avb3 induces structural changes associated with integrin activation. An engineered high-affinity FN10 mutant binds in a radically different orientation, explaining why it is not activating. 16. Zhu J, Zhu J, Springer TA: Complete integrin headpiece opening in eight steps. J Cell Biol 2013, 201:1053-1068. A series of crystal structures of integrin aIIbb3 with RGD ligand allows the activation pathway to be mapped in unprecedented detail. Headpiece opening is estimated to increase the affinity for ligand 200-fold. 17. Leitinger B: Transmembrane collagen receptors. Annu Rev Dev Cell Biol 2011, 27:265-290. 18. Emsley J, Knight CG, Farndale RW, Barnes MJ, Liddington RC: Structural basis of collagen recognition by integrin a2b1. Cell 2000, 101:47-56. 19. Sen M, Yuki K, Springer TA: An internal ligand-bound, metastable state of a leukocyte integrin, aXb2. J Cell Biol 2013, 203:629-642. This paper reports the first high-resolution structure of an integrin ectodomain with an a I domain. Although the overall structure is bent, the a I domain is in the active conformation with its C-terminal linker segment providing an internal ligand for the b I domain. The linkage of the a I domain to the propeller domain is unexpectedly loose, allowing a wide range of motions. 20. Alonso JL, Essafi M, Xiong JP, Stehle T, Arnaout MA: Does the integrin aA domain act as a ligand for its bA domain? Curr Biol 2002, 12:R340-R342. 21. Anthis NJ, Campbell ID: The tail of integrin activation. Trends Biochem Sci 2011, 36:191-198. 22. Calderwood DA, Campbell ID, Critchley DR: Talins and kindlins: partners in integrin-mediated adhesion. Nat Rev Mol Cell Biol 2013, 14:503-517. 23. Das M, Subbayya Ithychanda S, Qin J, Plow EF: Mechanisms of talin-dependent integrin signaling and crosstalk. Biochim Biophys Acta 2014, 1838:579-588. www.sciencedirect.com
24. Moser M, Legate KR, Zent R, Fassler R: The tail of integrins, talin, and kindlins. Science 2009, 324:895-899. 25. Shattil SJ, Kim C, Ginsberg MH: The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol 2010, 11:288-300. 26. Ye F, Petrich BG, Anekal P, Lefort CT, Kasirer-Friede A, Shattil SJ, Ruppert R, Moser M, Fassler R, Ginsberg MH: The mechanism of kindlin-mediated activation of integrin aIIbb3. Curr Biol 2013, 23:2288-2295. 27. Anthis NJ, Wegener KL, Ye F, Kim C, Goult BT, Lowe ED, Vakonakis I, Bate N, Critchley DR, Ginsberg MH, Campbell ID: The structure of an integrin/talin complex reveals the basis of inside-out signal transduction. EMBO J 2009, 28:3623-3632. 28. Garcia-Alvarez B, de Pereda JM, Calderwood DA, Ulmer TS, Critchley D, Campbell ID, Ginsberg MH, Liddington RC: Structural determinants of integrin recognition by talin. Mol Cell 2003, 11:49-58. 29. Yates LA, Fuzery AK, Bonet R, Campbell ID, Gilbert RJ: Biophysical analysis of kindlin-3 reveals an elongated conformation and maps integrin binding to the membranedistal b-subunit NPXY motif. J Biol Chem 2012, 287:37715-37731. 30. Wegener KL, Partridge AW, Han J, Pickford AR, Liddington RC, Ginsberg MH, Campbell ID: Structural basis of integrin activation by talin. Cell 2007, 128:171-182. 31. Goult BT, Zacharchenko T, Bate N, Tsang R, Hey F, Gingras AR, Elliott PR, Roberts GC, Ballestrem C, Critchley DR, Barsukov IL: RIAM and vinculin binding to talin are mutually exclusive and regulate adhesion assembly and turnover. J Biol Chem 2013, 288:8238-8249. A brute-force approach yields the structures of helical bundles R3-R6 of the talin and thereby completes the structural characterisation of the talin rod domain. 32. Izard T, Evans G, Borgon RA, Rush CL, Bricogne G, Bois PR: Vinculin activation by talin through helical bundle conversion. Nature 2004, 427:171-175. 33. Papagrigoriou E, Gingras AR, Barsukov IL, Bate N, Fillingham IJ, Patel B, Frank R, Ziegler WH, Roberts GC, Critchley DR, Emsley J: Activation of a vinculin-binding site in the talin rod involves rearrangement of a five-helix bundle. EMBO J 2004, 23:2942-2951. 34. del Rio A, Perez-Jimenez R, Liu R, Roca-Cusachs P, Fernandez JM, Sheetz MP: Stretching single talin rod molecules activates vinculin binding. Science 2009, 323:638-641. 35. Bakolitsa C, Cohen DM, Bankston LA, Bobkov AA, Cadwell GW, Jennings L, Critchley DR, Craig SW, Liddington RC: Structural basis for vinculin activation at sites of cell adhesion. Nature 2004, 430:583-586. 36. Ziegler WH, Liddington RC, Critchley DR: The structure and regulation of vinculin. Trends Cell Biol 2006, 16:453-460. 37. Goult BT, Xu XP, Gingras AR, Swift M, Patel B, Bate N, Kopp PM, Barsukov IL, Critchley DR, Volkmann N, Hanein D: Structural studies on full-length talin1 reveal a compact auto-inhibited dimer: implications for talin activation. J Struct Biol 2013, 184:21-32. This paper reports the structure at 25 A˚ resolution of full-length talin. The integrin-binding heads are sequestered inside a globular dimer, consistent with the view that talin needs to be activated for integrin binding. 38. Song X, Yang J, Hirbawi J, Ye S, Perera HD, Goksoy E, Dwivedi P, Plow EF, Zhang R, Qin J: A novel membrane-dependent on/off switch mechanism of talin ferm domain at sites of cell adhesion. Cell Res 2012, 22:1533-1545. The crystal structure of a talin head region bound to rod segment R9 reveals the mechanism of auto-inhibition. The authors propose an electrostatic ‘pull–push’ mechanism for talin activation at PIP2-enriched membranes. 39. Kanchanawong P, Shtengel G, Pasapera AM, Ramko EB, Davidson MW, Hess HF, Waterman CM: Nanoscale architecture of integrin-based cell adhesions. Nature 2010, 468:580-584. Current Opinion in Structural Biology 2014, 29:10–16
16 Multi-protein assemblies in signalling
40. Shrivastava A, Radziejewski C, Campbell E, Kovac L, McGlynn M, Ryan TE, Davis S, Goldfarb MP, Glass DJ, Lemke G, Yancopoulos GD: An orphan receptor tyrosine kinase family whose members serve as nonintegrin collagen receptors. Mol Cell 1997, 1:25-34. 41. Vogel W, Gish GD, Alves F, Pawson T: The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell 1997, 1:13-23. 42. Fu HL, Valiathan RR, Arkwright R, Sohail A, Mihai C, Kumarasiri M, Mahasenan KV, Mobashery S, Huang P, Agarwal G, Fridman R: Discoidin domain receptors: unique receptor tyrosine kinases in collagen-mediated signaling. J Biol Chem 2013, 288:74307437.
domain receptors: binding sites on collagens II and III and molecular determinants for collagen IV recognition by DDR1. Matrix Biol 2011, 30:16-26. 48. Carafoli F, Bihan D, Stathopoulos S, Konitsiotis AD, Kvansakul M, Farndale RW, Leitinger B, Hohenester E: Crystallographic insight into collagen recognition by discoidin domain receptor 2. Structure 2009, 17:1573-1581. 49. Ichikawa O, Osawa M, Nishida N, Goshima N, Nomura N, Shimada I: Structural basis of the collagen-binding mode of discoidin domain receptor 2. EMBO J 2007, 26:4168-4176. 50. Lemmon MA, Schlessinger J: Cell signaling by receptor tyrosine kinases. Cell 2010, 141:1117-1134.
43. Xu H, Bihan D, Chang F, Huang PH, Farndale RW, Leitinger B: Discoidin domain receptors promote a1b1- and a2b1-integrin mediated cell adhesion to collagen by enhancing integrin activation. PLoS ONE 2012, 7:e52209.
51. Ward CW, Menting JG, Lawrence MC: The insulin receptor changes conformation in unforeseen ways on ligand binding: sharpening the picture of insulin receptor activation. BioEssays 2013, 35:945-954.
44. Carafoli F, Mayer MC, Shiraishi K, Pecheva MA, Chan LY, Nan R, Leitinger B, Hohenester E: Structure of the discoidin domain receptor 1 extracellular region bound to an inhibitory Fab fragment reveals features important for signaling. Structure 2012, 20:688-697. This paper reports the first crystal structure of a complete DDR1 ectodomain. Mutagenesis and epitope mapping of function-blocking antibodies were used to identify regions required for transmembrane signalling.
52. Hubbard SR: Juxtamembrane autoinhibition in receptor tyrosine kinases. Nat Rev Mol Cell Biol 2004, 5:464-471.
45. Farndale RW, Lisman T, Bihan D, Hamaia S, Smerling CS, Pugh N, Konitsiotis A, Leitinger B, de Groot PG, Jarvis GE, Raynal N: Cell– collagen interactions: the use of peptide toolkits to investigate collagen–receptor interactions. Biochem Soc Trans 2008, 36:241-250. 46. Konitsiotis AD, Raynal N, Bihan D, Hohenester E, Farndale RW, Leitinger B: Characterization of high affinity binding motifs for the discoidin domain receptor DDR2 in collagen. J Biol Chem 2008, 283:6861-6868. 47. Xu H, Raynal N, Stathopoulos S, Myllyharju J, Farndale RW, Leitinger B: Collagen binding specificity of the discoidin
Current Opinion in Structural Biology 2014, 29:10–16
53. Jura N, Zhang X, Endres NF, Seeliger MA, Schindler T, Kuriyan J: Catalytic control in the EGF receptor and its connection to general kinase regulatory mechanisms. Mol Cell 2011, 42:9-22. 54. Noordeen NA, Carafoli F, Hohenester E, Horton MA, Leitinger B: A transmembrane leucine zipper is required for activation of the dimeric receptor tyrosine kinase DDR1. J Biol Chem 2006, 281:22744-22751. 55. Xu H, Abe T, Liu JK, Zalivina I, Hohenester E, Leitinger B: Normal activation of discoidin domain receptor 1 mutants with disulfide cross-links, insertions, or deletions in the extracellular juxtamembrane region: mechanistic implications. J Biol Chem 2014, 289:13565-13574. Extensive mutagenesis is used to show that the extracellular juxtamembrane region of DDR1 is unlikely to transmit a conformational change. The authors conclude that DDRs may become activated by collagen-induced clustering.
www.sciencedirect.com
ScienceDirect Signalling complexes at the cell-matrix interface Erhard Hohenester The extracellular matrix critically controls cell behaviour. Many cell-matrix interactions are mediated by transmembrane receptors of the integrin family. In the last two years, the structural changes resulting from ligand binding to integrins a5b1, avb3 and aIIbb3 have been mapped in unprecedented detail. The structure of integrin aXb2 has revealed how ligand binding to the a I domain is transmitted to the rest of the ectodomain. The structural characterisation of the cytosolic regulator talin has been continued, revealing how the integrin binding site is blocked in auto-inhibited talin. Finally, structures of the discoidin domain receptors DDR1 and DDR2 have begun to reveal how these atypical receptor tyrosine kinases become activated by the major matrix component collagen. Addresses Department of Life Sciences, Imperial College London, Sir Ernst Chain Building, London SW7 2AZ, UK Corresponding author: Hohenester, Erhard ([email protected])
Current Opinion in Structural Biology 2014, 29:10–16 This review comes from a themed issue on Multi-protein assemblies in signalling Edited by Deborah Fass and Katrin Rittinger
http://dx.doi.org/10.1016/j.sbi.2014.08.009 0959-440X/# 2014 Elsevier Ltd. All right reserved.
Introduction Most cells in higher animals are organised into tissues, which consist of differentiated cells in contact with each other and/or with extracellular matrix (ECM). The ECM is composed of large glycoproteins, which frequently are assembled into supramolecular structures [1,2]. Sheetlike matrices termed basement membranes cover the basal side of all epithelia and surround muscle, fat and peripheral nerve cells. Fibrillar matrices impart tensile strength (collagen fibres) and elasticity (elastic fibres) to many tissues. In addition to providing structural support for cells, the ECM controls many fundamental cellular processes, such as cell survival, polarisation, differentiation and migration. Some of these effects are achieved by modulating the diffusion and availability of growth factors, which signal through enzyme-linked transmembrane receptors. However, many signals derive from the composition and biomechanical properties of the ECM per se. Current Opinion in Structural Biology 2014, 29:10–16
Matrix stiffness is now recognised as a powerful regulator of cell differentiation, for instance, but how stiffness regulates gene expression remains incompletely understood [3]. A major hub of cell-matrix signalling are focal contacts and adhesions, in which transmembrane receptors of the integrin family link ECM molecules to a complex and dynamic network of intracellular adaptors and the cytoskeleton [4,5].
Integrin structure and shape-shifting The 24 mammalian integrins are assembled from 18 a subunits and eight b subunits, and include receptors for the major ECM proteins fibronectin (e.g. a5b1, avb3), laminins (e.g. a3b1, a6b1, a7b1) and collagens (e.g. a1b1, a2b1) [6]. Integrins have a large extracellular region consisting of a ligand binding head and two legs, which are connected to single transmembrane helices and short cytoplasmic domains (Figure 1a). The globular head is constructed from both subunits: the a subunit contributes a b-propeller domain, which in some integrins contains an inserted (a I) domain; the b subunit contributes the b I domain (so called because of its homology to the a I domain). When present, the a I domain contains the ligand binding site. In all other integrins, the ligand is believed to bind at the a/b interface with major determinants residing in the b I domain. Ligand binding involves metal ion-dependent adhesion sites (MIDAS) in both a and b I domains and triggers similar conformational transitions within the I domains (most importantly, a piston-like movement of the C-terminal helix a7) [7]. The first crystal structure of a complete integrin ectodomain (avb3) revealed a severely bent conformation that is now believed to represent a state with low affinity for ligand (Figure 1a) [8,9]. A dramatically different conformation with straight and separated legs was observed by electron microscopy and interpreted as the high-affinity state [10]. A pertinent feature of the high-affinity conformation, namely the swing-out of the b hybrid domain (Figure 1a), was observed in a crystal structure of the ligand-binding headpiece of platelet integrin aIIbb3 [11]. These and many other studies strongly support an allosteric mechanism of integrin activation [6,7], but the delicate energetic balance of integrin shape-shifting continues to throw up surprises (see below). Importantly, a recent single particle reconstruction of full-length aIIbb3 integrin embedded in nanodiscs suggest that the physiological resting state may differ substantially from the bent conformation observed in several crystal structures [12]. Given the demonstrated importance of the transmembrane region in integrin activation [13], further structural www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 11
studies of integrins in their native environment are clearly warranted. Several recent studies have shown that bent integrins in the crystalline state can bind ligand and display structural features typical of activated integrins. The structure of the a5b1 integrin headpiece (the first structure of a member of the large and ancestral b1 integrin subfamily) was determined in the low-affinity (closed) conformation in complex with an inhibitory allosteric antibody [14]. Similarly to the bent avb3 and aIIbb3 integrins, the closed a5b1 integrin headpiece nevertheless bound a fibronectin-mimetic Arg-Gly-Asp (RGD) peptide. This information allowed the in silico docking of the integrinbinding FN9-FN10 region of fibronectin, which helped identify the residues in the integrin a5 subunit that interact with the so-called synergy site in FN9 [14]. In another study, a crystal structure of avb3 integrin with fibronectin FN10 was obtained by soaking crystals of bent avb3 integrin with FN10 protein. FN10 binding was associated with structural changes in the b I domain, which were not observed with a purely antagonistic FN10 mutant [15]. Similarly, soaking of RGD peptides into crystals of the closed aIIbb3 headpiece allowed various intermediate states on the pathway to the highaffinity conformation to be captured, including a structure with a partially swung out b3 subunit hybrid domain [16]. The collagen-binding integrins a1b1, a2b1, a10b1 and a11b1 all contain an a I domain [17] that responds to collagen binding with a downward movement of the Cterminal a7 helix [18]. How this change is transmitted to the remainder of the integrin ectodomain, and ultimately the cytosolic domains, has been unclear. A structure of the leucocyte integrin aXb2 has provided the first detailed view of the activation mechanism of a I domain-containing integrins [19]. Even though the overall conformation of aXb2 integrin was bent, the a I domain was captured in the high-affinity conformation with the a7 helix in the downward position (Figure 1b). The segment following the a7 helix contains a conserved glutamic acid (Glu318 in aXb2) that had previously been hypothesised to provide an internal ligand for the b I domain [20]. In the new crystal structure, Glu318 indeed coordinates the MIDAS magnesium ion in the b I domain, thus priming the transition to the active conformation of the b I domain and the eventual swing-out of the hybrid domain [19].
Cytoplasmic interactions with integrins The conformational changes in the integrin heterodimer are used to transmit signals across the plasma membrane in both directions [6,7]. In outside-in signalling, ligand binding to the ectodomain leads to a separation of the two legs, which allows the b subunit cytoplasmic domain to bind intracellular proteins such as talin and kindlins. In inside-out signalling, intracellular activation of talin www.sciencedirect.com
(discussed below) leads to its recruitment to the plasma membrane, where it can bind to the b subunit cytoplasmic domain, thereby triggering the transition of the integrin heterodimer to a state with high affinity for extracellular ligand. How talin binds to the integrin b subunit and activates integrins is now well understood, but a molecular explanation for kindlins’ essential roles in integrin signalling is currently lacking [21–25]. A recent study has shown that kindlins are required for the clustering of talin-activated integrin [26]. Talin and kindlins have in common an N-terminal atypical FERM (4.1 protein, ezrin, radixin, moesin) domain that mediates integrin binding [27–30]. In talin (but not in kindlins) this so-called head domain is followed by a long rod consisting of 13 a-helix bundles and terminating in a dimerisation domain (Figure 2a) [22]. The structural characterisation of the rod was recently completed in a tour de force, in which NMR spectroscopy was used to define folded units before detailed structure determination [31]. The a-helix bundles of talin contain multiple binding sites for the actin-binding protein vinculin. Vinculin binding is mediated by the hydrophobic surfaces of amphipathic helices in talin, which become exposed in a process of a-helix bundle conversion [32,33]. A singlemolecule study has shown that mechanical stretching of the talin rod activates vinculin binding, providing a mechanism for force-sensing at focal adhesions [34]. Vinculin also needs to be restructured in order to bind talin, but there is disagreement about the precise mechanism of vinculin activation [32,35,36]. In contrast to the isolated talin rod, in which the flexibly linked a-helix bundles form an extended array, fulllength (dimeric) talin assumes a compact conformation that represents an autoinhibited form with regard to integrin binding. A recent single particle reconstruction of full-length talin places the two heads of the dimer inside a doughnut-shaped structure formed by the rods [37]. A key feature of this assembly is the interaction of the F3 subdomain of the talin head with the ninth a-helix bundle, R9. The crystal structure of a talin F2F3-R9 complex showed that the rod segment not only occludes the integrin binding site in F3, but also antagonises a positively charged surface in F2 that mediates the interaction of the talin head with the plasma membrane (Figure 2b,c) [38]. While the reductionist approach of traditional structural investigations has been very successful in defining the key interactions in integrin-mediated cell-matrix contacts, a complete picture will only emerge from structural studies of focal adhesions in situ [4]. Super-resolution fluorescence microscopy has been used to determine the vertical organisation of focal adhesions formed by cells plated on fibronectin-coated glass [39]. The layer closest to the plasma membrane contains the integrin Current Opinion in Structural Biology 2014, 29:10–16
12 Multi-protein assemblies in signalling
Figure 1
Ligand
(a) Ligand
α I domain Conserved Glu
Propeller domain
β I domain Hybrid domain
Thigh domain
α
α
β
Closed headpiece, bent
PSI domain
α
β
α
β
Open headpiece, extended
Closed headpiece, extended
β
α I domain-containing integrin
MIDAS
(b)
αI
βI
Propeller
Thigh
Hybrid
PSI
α
β Current Opinion in Structural Biology
Integrin structure and conformational changes. (a) Schematic drawings of integrins in different functional states (adapted from [7]). The coloured domains represent actual crystal structures (except for the situation on the right), whereas the locations of the grey domains are inferred from electron microscopic reconstructions. The plasma membrane is represented by a grey bar. During headpiece opening, a downward movement of the Cterminal helix of the b I domain (magenta) is coupled to a swing-out of the hybrid and plexin-semaphorin-integrin (PSI) domains. In integrins with an a I domain, a similar movement of the C-terminal helix in the a I domain (red) results in internal ligation of the b I domain metal ion-dependent adhesion site (MIDAS), which in turn leads to hybrid domain swing-out. (b) Crystal structure of integrin aXb2 in a metastable conformation [19]. Only the headpiece portion is shown for clarity. Calcium and magnesium ions are shown as black and yellow spheres, respectively. The a I domain is in the active conformation, with the C-terminal helix in the down position and the following segment (red) ligating the b I domain MIDAS. The C-terminal helix of the b I domain (magenta) is trapped in an intermediate position and the headpiece is closed. The positions of the hybrid and PSI domains in the open conformation, as observed in an integrin aIIbb3 structure [11], are indicated in grey.
Current Opinion in Structural Biology 2014, 29:10–16
www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 13
Figure 2
(a)
Head
Rod R8
R4 R2 F0 F1 F2 F3
R5
R6
R7
R9
R10
R11
R12
R13
DD
R1 R3
(b)
(c) Negative patch + + + +
Integrin β1 cytoplasmic region
R9 F2 F3
Current Opinion in Structural Biology
Talin structure and autoinhibition. (a) Domain structure of talin. The talin head consists of a FERM (4.1 protein, ezrin, radixin, moesin) domain with linearly arranged subdomains F0–F4. The talin rod consists of 13 a-helix bundles [31]. R1-R4 and R8 are 4-helix bundles, which have their N-termini and C-termini close together, whereas the remaining bundles have five helices, allowing a linear head-to-tail arrangement. Full-length talin is dimerised via the C-terminal dimerisation domain (DD) and adopts a compact autoinhibited conformation [37] that is converted to a more open conformation by a variety of activators. The talin rod contains multiple binding sites for the actin-binding protein vinculin (marked by yellow stars) that become exposed by mechanical tension. (b) Crystal structure of the talin-2 F2F3 domains bound to the cytoplasmic region of integrin b1 [27]. The plasma membrane is represented by a grey bar. A positive patch on the F2 domain interacts with the phospholipid headgroups. (c) Crystal structure of the talin-1 F2F3 domains bound to the rod segment R9, representing the autoinhibited state of talin [38]. R9 blocks the integrin binding site and contains a negative patch that disfavours membrane binding.
cytoplasmic domains, the talin head domain, paxillin and focal adhesion kinase. This is followed by a ‘force transduction layer’ composed of the talin rod domain and vinculin. Actin was detected only at greater than 40 nm distance from the plasma membrane [39].
and sustained. The DDRs do not mediate firm cell adhesion themselves, but they influence adhesion indirectly by regulating the activation state of collagen-binding integrins [43]. However, in common with typical RTKs, the main functions of DDRs appear to be related to signalling and the control of cell proliferation [17,42].
Discoidin domain receptors Integrins are not the only receptors for ECM molecules. A particularly intriguing class of non-integrin ECM receptors are the two discoidin domain receptors, DDR1 and DDR2, which are receptor tyrosine kinases (RTKs) that are activated by collagen [17,40–42]. Compared with typical RTKs that are activated by small diffusible growth factors, DDR activation by collagens is slow www.sciencedirect.com
DDRs consist of an N-terminal discoidin (DS) domain that contains the collagen binding site, followed by a DSlike domain [44], a long extracellular juxtamembrane (JM) region (30–50 residues), a single transmembrane helix, an even longer cytoplasmic JM region (up to 160 residues), and a canonical tyrosine kinase domain. Systematic investigations using collagen peptide libraries Current Opinion in Structural Biology 2014, 29:10–16
14 Multi-protein assemblies in signalling
identified a GVMGFO motif (O is hydroxyproline) as the major DDR binding site in fibrillar collagens; this site is distinct from all known integrin binding sites [45–47]. Triple-helical peptides containing the GVMGFO motif bind to an amphipathic pocket at the top of the DS domain of DDRs (Figure 3) [48,49]. How collagen binding leads to DDR activation (i.e. autophosphorylation of the cytoplasmic regions) is still unclear. Typical RTKs respond to ligand binding by dimerisation or by structural changes within a constitutive dimer [50,51]. In either case, the JM regions frequently play an important role in activation by restraining the inactive RTK and by transmitting structural changes across the plasma membrane [52,53]. DDRs are constitutive noncovalent dimers [54]. A recent study showed that the extracellular JM region of DDR1 is extremely tolerant to mutation: DDR1 activation was unaffected by disulphide cross-links, deletions or insertions of flexible segments [55]. These findings argue against a mechanism of conformational coupling across the plasma membrane and are Figure 3
more consistent with signalling by DDR clusters, but information about DDR stoichiometry at the cell surface is currently lacking.
Concluding remarks Integrin-mediated adhesion and signalling is now quite well understood at the level of single integrin heterodimers. The challenge for the future is to integrate these insights with structural and functional studies using cultured cells and observations from animal experiments. One potential concern is that the prominent focal adhesions observed in cells plated on stiff two-dimensional substrates may not always be representative of the situation in three-dimensional matrices, which are less amenable to detailed structural analyses. Another interesting question is how well the ECM substrates that are used in vitro actually correspond to the situation in vivo. This is particularly troubling in the case of collagen binding by integrins and DDRs. Almost all experiments have been performed with isolated triple helices, but many collagens form highly organised fibrils, the surface of which is decorated with a variety of accessory ECM proteins. It will, therefore, be important to investigate how integrins and DDRs interact with native fibrillar collagens.
Collagen
Conflict of interest None declared.
Acknowledgements This review is dedicated to the memory of Iain D Campbell, who pioneered structural studies of integrin activation. I thank Birgit Leitinger and David Pulido for critical reading of the manuscript. EH is supported by the Wellcome Trust.
DS domain
References and recommended reading Papers of particular interest, published within the period of review, have been highlighted as: of special interest of outstanding interest
DS-like domain
JM region (50 aa)
Current Opinion in Structural Biology
Discoidin domain receptor structure. The figure shows a composite structure, consisting of the unliganded DDR1 ectodomain [44] and a collagen triple helix (orange) docked according to the crystal structure of the DDR2 discoidin (DS) domain bound to a collagen-like peptide [48]. A surface patch on the DS domain that is essential for DDR activation [44] is coloured magenta. Disulphide bridges and N-linked glycans are in yellow and light pink, respectively. A flexible 50-residue juxtamembrane (JM) region links the DS-like domain to the transmembrane region, which mediates the constitutive dimerisation of DDRs [54,55]. Current Opinion in Structural Biology 2014, 29:10–16
1.
Hynes RO: The extracellular matrix: not just pretty fibrils. Science 2009, 326:1216-1219.
2.
Bruckner P: Suprastructures of extracellular matrices: paradigms of functions controlled by aggregates rather than molecules. Cell Tissue Res 2010, 339:7-18.
3.
Discher DE, Janmey P, Wang YL: Tissue cells feel and respond to the stiffness of their substrate. Science 2005, 310:1139-1143.
4.
Hanein D, Horwitz AR: The structure of cell-matrix adhesions: the new frontier. Curr Opin Cell Biol 2012, 24:134-140.
5.
Wolfenson H, Lavelin I, Geiger B: Dynamic regulation of the structure and functions of integrin adhesions. Dev Cell 2013, 24:447-458.
6.
Hynes RO: Integrins: bidirectional, allosteric signaling machines. Cell 2002, 110:673-687.
7.
Luo BH, Carman CV, Springer TA: Structural basis of integrin regulation and signaling. Annu Rev Immunol 2007, 25:619-647.
8.
Xiong JP, Stehle T, Diefenbach B, Zhang R, Dunker R, Scott DL, Joachimiak A, Goodman SL, Arnaout MA: Crystal structure of the extracellular segment of integrin avb3. Science 2001, 294:339-345. www.sciencedirect.com
Signalling complexes at the cell-matrix interface Hohenester 15
9.
Xiong JP, Stehle T, Zhang R, Joachimiak A, Frech M, Goodman SL, Arnaout MA: Crystal structure of the extracellular segment of integrin avb3 in complex with an Arg-Gly-Asp ligand. Science 2002, 296:151-155.
10. Takagi J, Petre BM, Walz T, Springer TA: Global conformational rearrangements in integrin extracellular domains in outside-in and inside-out signaling. Cell 2002, 110:511-599. 11. Xiao T, Takagi J, Coller BS, Wang JH, Springer TA: Structural basis for allostery in integrins and binding to fibrinogenmimetic therapeutics. Nature 2004, 432:59-67. 12. Choi WS, Rice WJ, Stokes DL, Coller BS: Three-dimensional reconstruction of intact human integrin aIIbb3: new implications for activation-dependent ligand binding. Blood 2013, 122:4165-4171. This paper reports the structure at 20 A˚ resolution of integrin aIIbb3 embedded in a nanodisc lipid bilayer. The ligand-binding headpiece in the low-affinity state is more accessible than predicted from crystal structures and the arrangement of the two legs is also different, leading the authors to propose modification of the switchblade model of integrin extension [7]. 13. Kim C, Schmidt T, Cho EG, Ye F, Ulmer TS, Ginsberg MH: Basic amino-acid side chains regulate transmembrane integrin signalling. Nature 2012, 481:209-213. This study shows that a ‘snorkelling’ lysine in the transmembrane region of the integrin b3 domain is required for correct membrane embedding. Mutation of this lysine causes integrin activation by changing the membrane-crossing angle. 14. Nagae M, Re S, Mihara E, Nogi T, Sugita Y, Takagi J: Crystal structure of a5b1 integrin ectodomain: atomic details of the fibronectin receptor. J Cell Biol 2012, 197:131-140. This paper reports the first structure of an integrin from the b1 subfamily. Modelling shows how integrin a5b1 binds fibronectin through a primary interaction with the RGD sequence in FN10 and a synergy site in FN9. 15. Van Agthoven JF, Xiong JP, Alonso JL, Rui X, Adair BD, Goodman SL, Arnaout MA: Structural basis for pure antagonism of integrin avb3 by a high-affinity form of fibronectin. Nat Struct Mol Biol 2014, 21:383-388. This paper reports that fibronectin FN10 binding to integrin avb3 induces structural changes associated with integrin activation. An engineered high-affinity FN10 mutant binds in a radically different orientation, explaining why it is not activating. 16. Zhu J, Zhu J, Springer TA: Complete integrin headpiece opening in eight steps. J Cell Biol 2013, 201:1053-1068. A series of crystal structures of integrin aIIbb3 with RGD ligand allows the activation pathway to be mapped in unprecedented detail. Headpiece opening is estimated to increase the affinity for ligand 200-fold. 17. Leitinger B: Transmembrane collagen receptors. Annu Rev Dev Cell Biol 2011, 27:265-290. 18. Emsley J, Knight CG, Farndale RW, Barnes MJ, Liddington RC: Structural basis of collagen recognition by integrin a2b1. Cell 2000, 101:47-56. 19. Sen M, Yuki K, Springer TA: An internal ligand-bound, metastable state of a leukocyte integrin, aXb2. J Cell Biol 2013, 203:629-642. This paper reports the first high-resolution structure of an integrin ectodomain with an a I domain. Although the overall structure is bent, the a I domain is in the active conformation with its C-terminal linker segment providing an internal ligand for the b I domain. The linkage of the a I domain to the propeller domain is unexpectedly loose, allowing a wide range of motions. 20. Alonso JL, Essafi M, Xiong JP, Stehle T, Arnaout MA: Does the integrin aA domain act as a ligand for its bA domain? Curr Biol 2002, 12:R340-R342. 21. Anthis NJ, Campbell ID: The tail of integrin activation. Trends Biochem Sci 2011, 36:191-198. 22. Calderwood DA, Campbell ID, Critchley DR: Talins and kindlins: partners in integrin-mediated adhesion. Nat Rev Mol Cell Biol 2013, 14:503-517. 23. Das M, Subbayya Ithychanda S, Qin J, Plow EF: Mechanisms of talin-dependent integrin signaling and crosstalk. Biochim Biophys Acta 2014, 1838:579-588. www.sciencedirect.com
24. Moser M, Legate KR, Zent R, Fassler R: The tail of integrins, talin, and kindlins. Science 2009, 324:895-899. 25. Shattil SJ, Kim C, Ginsberg MH: The final steps of integrin activation: the end game. Nat Rev Mol Cell Biol 2010, 11:288-300. 26. Ye F, Petrich BG, Anekal P, Lefort CT, Kasirer-Friede A, Shattil SJ, Ruppert R, Moser M, Fassler R, Ginsberg MH: The mechanism of kindlin-mediated activation of integrin aIIbb3. Curr Biol 2013, 23:2288-2295. 27. Anthis NJ, Wegener KL, Ye F, Kim C, Goult BT, Lowe ED, Vakonakis I, Bate N, Critchley DR, Ginsberg MH, Campbell ID: The structure of an integrin/talin complex reveals the basis of inside-out signal transduction. EMBO J 2009, 28:3623-3632. 28. Garcia-Alvarez B, de Pereda JM, Calderwood DA, Ulmer TS, Critchley D, Campbell ID, Ginsberg MH, Liddington RC: Structural determinants of integrin recognition by talin. Mol Cell 2003, 11:49-58. 29. Yates LA, Fuzery AK, Bonet R, Campbell ID, Gilbert RJ: Biophysical analysis of kindlin-3 reveals an elongated conformation and maps integrin binding to the membranedistal b-subunit NPXY motif. J Biol Chem 2012, 287:37715-37731. 30. Wegener KL, Partridge AW, Han J, Pickford AR, Liddington RC, Ginsberg MH, Campbell ID: Structural basis of integrin activation by talin. Cell 2007, 128:171-182. 31. Goult BT, Zacharchenko T, Bate N, Tsang R, Hey F, Gingras AR, Elliott PR, Roberts GC, Ballestrem C, Critchley DR, Barsukov IL: RIAM and vinculin binding to talin are mutually exclusive and regulate adhesion assembly and turnover. J Biol Chem 2013, 288:8238-8249. A brute-force approach yields the structures of helical bundles R3-R6 of the talin and thereby completes the structural characterisation of the talin rod domain. 32. Izard T, Evans G, Borgon RA, Rush CL, Bricogne G, Bois PR: Vinculin activation by talin through helical bundle conversion. Nature 2004, 427:171-175. 33. Papagrigoriou E, Gingras AR, Barsukov IL, Bate N, Fillingham IJ, Patel B, Frank R, Ziegler WH, Roberts GC, Critchley DR, Emsley J: Activation of a vinculin-binding site in the talin rod involves rearrangement of a five-helix bundle. EMBO J 2004, 23:2942-2951. 34. del Rio A, Perez-Jimenez R, Liu R, Roca-Cusachs P, Fernandez JM, Sheetz MP: Stretching single talin rod molecules activates vinculin binding. Science 2009, 323:638-641. 35. Bakolitsa C, Cohen DM, Bankston LA, Bobkov AA, Cadwell GW, Jennings L, Critchley DR, Craig SW, Liddington RC: Structural basis for vinculin activation at sites of cell adhesion. Nature 2004, 430:583-586. 36. Ziegler WH, Liddington RC, Critchley DR: The structure and regulation of vinculin. Trends Cell Biol 2006, 16:453-460. 37. Goult BT, Xu XP, Gingras AR, Swift M, Patel B, Bate N, Kopp PM, Barsukov IL, Critchley DR, Volkmann N, Hanein D: Structural studies on full-length talin1 reveal a compact auto-inhibited dimer: implications for talin activation. J Struct Biol 2013, 184:21-32. This paper reports the structure at 25 A˚ resolution of full-length talin. The integrin-binding heads are sequestered inside a globular dimer, consistent with the view that talin needs to be activated for integrin binding. 38. Song X, Yang J, Hirbawi J, Ye S, Perera HD, Goksoy E, Dwivedi P, Plow EF, Zhang R, Qin J: A novel membrane-dependent on/off switch mechanism of talin ferm domain at sites of cell adhesion. Cell Res 2012, 22:1533-1545. The crystal structure of a talin head region bound to rod segment R9 reveals the mechanism of auto-inhibition. The authors propose an electrostatic ‘pull–push’ mechanism for talin activation at PIP2-enriched membranes. 39. Kanchanawong P, Shtengel G, Pasapera AM, Ramko EB, Davidson MW, Hess HF, Waterman CM: Nanoscale architecture of integrin-based cell adhesions. Nature 2010, 468:580-584. Current Opinion in Structural Biology 2014, 29:10–16
16 Multi-protein assemblies in signalling
40. Shrivastava A, Radziejewski C, Campbell E, Kovac L, McGlynn M, Ryan TE, Davis S, Goldfarb MP, Glass DJ, Lemke G, Yancopoulos GD: An orphan receptor tyrosine kinase family whose members serve as nonintegrin collagen receptors. Mol Cell 1997, 1:25-34. 41. Vogel W, Gish GD, Alves F, Pawson T: The discoidin domain receptor tyrosine kinases are activated by collagen. Mol Cell 1997, 1:13-23. 42. Fu HL, Valiathan RR, Arkwright R, Sohail A, Mihai C, Kumarasiri M, Mahasenan KV, Mobashery S, Huang P, Agarwal G, Fridman R: Discoidin domain receptors: unique receptor tyrosine kinases in collagen-mediated signaling. J Biol Chem 2013, 288:74307437.
domain receptors: binding sites on collagens II and III and molecular determinants for collagen IV recognition by DDR1. Matrix Biol 2011, 30:16-26. 48. Carafoli F, Bihan D, Stathopoulos S, Konitsiotis AD, Kvansakul M, Farndale RW, Leitinger B, Hohenester E: Crystallographic insight into collagen recognition by discoidin domain receptor 2. Structure 2009, 17:1573-1581. 49. Ichikawa O, Osawa M, Nishida N, Goshima N, Nomura N, Shimada I: Structural basis of the collagen-binding mode of discoidin domain receptor 2. EMBO J 2007, 26:4168-4176. 50. Lemmon MA, Schlessinger J: Cell signaling by receptor tyrosine kinases. Cell 2010, 141:1117-1134.
43. Xu H, Bihan D, Chang F, Huang PH, Farndale RW, Leitinger B: Discoidin domain receptors promote a1b1- and a2b1-integrin mediated cell adhesion to collagen by enhancing integrin activation. PLoS ONE 2012, 7:e52209.
51. Ward CW, Menting JG, Lawrence MC: The insulin receptor changes conformation in unforeseen ways on ligand binding: sharpening the picture of insulin receptor activation. BioEssays 2013, 35:945-954.
44. Carafoli F, Mayer MC, Shiraishi K, Pecheva MA, Chan LY, Nan R, Leitinger B, Hohenester E: Structure of the discoidin domain receptor 1 extracellular region bound to an inhibitory Fab fragment reveals features important for signaling. Structure 2012, 20:688-697. This paper reports the first crystal structure of a complete DDR1 ectodomain. Mutagenesis and epitope mapping of function-blocking antibodies were used to identify regions required for transmembrane signalling.
52. Hubbard SR: Juxtamembrane autoinhibition in receptor tyrosine kinases. Nat Rev Mol Cell Biol 2004, 5:464-471.
45. Farndale RW, Lisman T, Bihan D, Hamaia S, Smerling CS, Pugh N, Konitsiotis A, Leitinger B, de Groot PG, Jarvis GE, Raynal N: Cell– collagen interactions: the use of peptide toolkits to investigate collagen–receptor interactions. Biochem Soc Trans 2008, 36:241-250. 46. Konitsiotis AD, Raynal N, Bihan D, Hohenester E, Farndale RW, Leitinger B: Characterization of high affinity binding motifs for the discoidin domain receptor DDR2 in collagen. J Biol Chem 2008, 283:6861-6868. 47. Xu H, Raynal N, Stathopoulos S, Myllyharju J, Farndale RW, Leitinger B: Collagen binding specificity of the discoidin
Current Opinion in Structural Biology 2014, 29:10–16
53. Jura N, Zhang X, Endres NF, Seeliger MA, Schindler T, Kuriyan J: Catalytic control in the EGF receptor and its connection to general kinase regulatory mechanisms. Mol Cell 2011, 42:9-22. 54. Noordeen NA, Carafoli F, Hohenester E, Horton MA, Leitinger B: A transmembrane leucine zipper is required for activation of the dimeric receptor tyrosine kinase DDR1. J Biol Chem 2006, 281:22744-22751. 55. Xu H, Abe T, Liu JK, Zalivina I, Hohenester E, Leitinger B: Normal activation of discoidin domain receptor 1 mutants with disulfide cross-links, insertions, or deletions in the extracellular juxtamembrane region: mechanistic implications. J Biol Chem 2014, 289:13565-13574. Extensive mutagenesis is used to show that the extracellular juxtamembrane region of DDR1 is unlikely to transmit a conformational change. The authors conclude that DDRs may become activated by collagen-induced clustering.
www.sciencedirect.com
