Structure of the Ire1 autophosphorylation complex and implications for the unfolded protein response
2011; Springer Nature; Volume: 30; Issue: 5 Linguagem: Inglês
10.1038/emboj.2011.18
ISSN1460-2075
AutoresMaruf M. U. Ali, Tina Bagratuni, Emma L. Davenport, Piotr Nowak, M. Cris Silva-Santisteban, Anthea Hardcastle, Craig McAndrews, Martin Rowlands, Gareth J. Morgan, Wynne Aherne, Ian Collins, Faith E. Davies, Laurence H. Pearl,
Tópico(s)Virus-based gene therapy research
ResumoArticle11 February 2011Open Access Structure of the Ire1 autophosphorylation complex and implications for the unfolded protein response Maruf M U Ali Corresponding Author Maruf M U Ali Section of Structural Biology, The Institute of Cancer Research, Chester Beatty Laboratories, London, UK Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London, UKThese authors are joint senior authors Search for more papers by this author Tina Bagratuni Tina Bagratuni Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Emma L Davenport Emma L Davenport Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Piotr R Nowak Piotr R Nowak Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London, UK Search for more papers by this author M Cris Silva-Santisteban M Cris Silva-Santisteban Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Anthea Hardcastle Anthea Hardcastle Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Craig McAndrews Craig McAndrews Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Martin G Rowlands Martin G Rowlands Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Gareth J Morgan Gareth J Morgan Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Wynne Aherne Wynne Aherne Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Ian Collins Ian Collins Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Faith E Davies Corresponding Author Faith E Davies Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UKThese authors are joint senior authors Search for more papers by this author Laurence H Pearl Corresponding Author Laurence H Pearl Section of Structural Biology, The Institute of Cancer Research, Chester Beatty Laboratories, London, UKThese authors are joint senior authorsPresent address: Genome Damage and Stability Centre, University of Sussex, Falmer, Brighton BN1 9RQ, UK Search for more papers by this author Maruf M U Ali Corresponding Author Maruf M U Ali Section of Structural Biology, The Institute of Cancer Research, Chester Beatty Laboratories, London, UK Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London, UKThese authors are joint senior authors Search for more papers by this author Tina Bagratuni Tina Bagratuni Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Emma L Davenport Emma L Davenport Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Piotr R Nowak Piotr R Nowak Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London, UK Search for more papers by this author M Cris Silva-Santisteban M Cris Silva-Santisteban Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Anthea Hardcastle Anthea Hardcastle Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Craig McAndrews Craig McAndrews Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Martin G Rowlands Martin G Rowlands Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Gareth J Morgan Gareth J Morgan Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK Search for more papers by this author Wynne Aherne Wynne Aherne Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Ian Collins Ian Collins Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK Search for more papers by this author Faith E Davies Corresponding Author Faith E Davies Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UKThese authors are joint senior authors Search for more papers by this author Laurence H Pearl Corresponding Author Laurence H Pearl Section of Structural Biology, The Institute of Cancer Research, Chester Beatty Laboratories, London, UKThese authors are joint senior authorsPresent address: Genome Damage and Stability Centre, University of Sussex, Falmer, Brighton BN1 9RQ, UK Search for more papers by this author Author Information Maruf M U Ali 1,2, Tina Bagratuni3, Emma L Davenport3, Piotr R Nowak2, M Cris Silva-Santisteban4, Anthea Hardcastle4, Craig McAndrews4, Martin G Rowlands4, Gareth J Morgan3, Wynne Aherne4, Ian Collins4, Faith E Davies 3 and Laurence H Pearl 1 1Section of Structural Biology, The Institute of Cancer Research, Chester Beatty Laboratories, London, UK 2Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London, UK 3Section of Haemato-Oncology, The Institute of Cancer Research, Surrey, UK 4Cancer Research UK Cancer Therapeutics Unit, The Institute of Cancer Research, Haddow Laboratories, Surrey, UK *Corresponding authors: Division of Molecular Biosciences, Department of Life Sciences, Imperial College London, London SW7 2AZ, UK. Tel.: +44 207 594 5733; E-mail: [email protected] of Haemato-Oncology, The Institute of Cancer Research, Brookes-Lawley Building, 15 Cotswold Road, Belmont, Sutton, Surrey SM2 5NG, UK. Tel.: +44 208 661 3670; Fax: +44 208 642 9634; E-mail: [email protected] Damage and Stability Centre, University of Sussex, Falmer, Brighton BN1 9RQ, UK. Tel.: +44 127 387 6544/Extn. 2699; Fax: +44 207 153 5457; E-mail: [email protected] The EMBO Journal (2011)30:894-905https://doi.org/10.1038/emboj.2011.18 Present address: Genome Damage and Stability Centre, University of Sussex, Falmer, Brighton BN1 9RQ, UK PDFDownload PDF of article text and main figures. Peer ReviewDownload a summary of the editorial decision process including editorial decision letters, reviewer comments and author responses to feedback. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions Figures & Info Ire1 (Ern1) is an unusual transmembrane protein kinase essential for the endoplasmic reticulum (ER) unfolded protein response (UPR). Activation of Ire1 by association of its N-terminal ER luminal domains promotes autophosphorylation by its cytoplasmic kinase domain, leading to activation of the C-terminal ribonuclease domain, which splices Xbp1 mRNA generating an active Xbp1s transcriptional activator. We have determined the crystal structure of the cytoplasmic portion of dephosphorylated human Ire1α bound to ADP, revealing the ‘phosphoryl-transfer’ competent dimeric face-to-face complex, which precedes and is distinct from the back-to-back RNase ‘active’ conformation described for yeast Ire1. We show that the Xbp1-specific ribonuclease activity depends on autophosphorylation, and that ATP-competitive inhibitors staurosporin and sunitinib, which inhibit autophosphorylation in vitro, also inhibit Xbp1 splicing in vivo. Furthermore, we demonstrate that activated Ire1α is a competent protein kinase, able to phosphorylate a heterologous peptide substrate. These studies identify human Ire1α as a target for development of ATP-competitive inhibitors that will modulate the UPR in human cells, which has particular relevance for myeloma and other secretory malignancies. Introduction The endoplasmic reticulum (ER) is a major compartment within the eukaryotic cell responsible for folding of secretory proteins. Disruption of the balance between secretory protein synthesis and the folding capacity of the ER activates a signalling network called the unfolded protein response (UPR). This attempts to minimise ER overload by transcriptional upregulation of ER chaperones, attenuation of translation, and activation of ER-associated degradation (ERAD) pathways (Ron and Walter, 2007; Todd et al, 2008; Zhang and Kaufman, 2008). If the effort to rectify the imbalance is not successful, the UPR switches from being pro-survival to eliciting an apoptotic response (Groenendyk and Michalak, 2005). Ire1 (a.k.a Ern1) is one of the three transmembrane ‘sensor’ proteins that propagate the signal from the lumen of the ER to the cytosol. Ire1 consists of an N-terminal luminal domain, a single-pass transmembrane spanning segment, and a cytosolic region subdivided into a Ser/Thr protein kinase domain and a C-terminal endoribonuclease (RNase) domain with homology to RNaseL (Tirasophon et al, 1998). The ER luminal domain of Ire1 is maintained in its inactive monomeric state by binding to the ER Hsp70 chaperone Bip. Accumulation of misfolded protein in the ER is thought to cause release of Bip (Bertolotti et al, 2000), allowing the luminal domain to dimerise (Credle et al, 2005; Zhou et al, 2006). This brings associated cytoplasmic domains into close proximity on the other side of the ER membrane, facilitating transautophosphorylation of the kinase domain, which results in activation of the RNase domain (Ron and Walter, 2007; Zhang and Kaufman, 2008). In budding yeast, Ire1 RNase activity is specific for Hac1 mRNA and excises a 252-nucleotide segment in a spliceosome-independent manner (Sidrauski and Walter, 1997; Gonzalez et al, 1999). Human Ire1 excises 26 nucleotides from Xbp1 mRNA. The spliced transcript Xbp1s encodes a potent transcriptional activator protein that upregulates expression of UPR target genes (Yoshida et al, 2001), including ER chaperones and components of the ERAD pathway (Friedlander et al, 2000). The UPR is activated in many human tumours and is a mechanism tumour cells exploit to survive hypoxia (Koumenis, 2006). The Ire1-Xbp1 branch of signalling is particularly important in multiple myeloma; a cancer resulting from malignant transformation of plasma cells. The survival rate for this cancer is low and thus far there is no direct curative therapy. Alongside its role in the UPR, Xbp1 controls plasma cell differentiation, and when misregulated, causes uncontrollable proliferation in myeloma cells (Davies et al, 2003; Iwakoshi et al, 2003; Carrasco et al, 2007; Davenport et al, 2007; Todd et al, 2008). As Xbp1s mRNA levels are directly dependent upon splicing by Ire1, preventing splicing of Xbp1 mRNA by pharmacological inhibition of the kinase/RNase activity of Ire1 would appear to be a very good strategy for selectively targeting multiple myeloma. However, the lack of a clear mechanistic understanding of Ire1 function, and in particular the coupling of the kinase and RNase activities, has clouded the issue. Early work in the field indicated that Ire1 acts like a typical receptor protein kinase, undergoing autophosphorylation in trans, following dimerisation of its ER luminal signalling domain (Shamu and Walter, 1996; Welihinda and Kaufman, 1996). Furthermore, a kinase-dead Ire1 was unable to splice substrate RNA (Shamu et al, 1994; Tirasophon et al, 1998), demonstrating that the Ire1 phosphoryl-transfer reaction was essential for RNase activation. However, the remarkable finding that binding of an ATP-competitive inhibitor to an active site mutant of yeast Ire1 resulted in RNase activation rather than inhibition has suggested that it is a conformational change in the kinase domain induced by nucleotide binding, rather than phosphorylation, that provides the trigger to activate the RNase (Papa et al, 2003). This phenomenon has also been reported for wild-type yeast Ire1 with the FDA-approved anti-cancer drug sunitinib (Korennykh et al, 2009). Recently, there have been several structures of the C-terminal kinase-RNase domains of yeast Ire1. The first of these structures (Lee et al, 2008) revealed a symmetric Ire1 dimer arranged in a back-to-back orientation and reported to be in an RNase ‘active’ conformation, but with the kinase active sites facing outwards in a manner that would preclude transphosphorylation. The biological significance of this arrangement is supported by mutations in the back-to-back interface that impair RNase activity. A second structure of yeast Ire1 has been described (Korennykh et al, 2009), in which the same back-to-back dimer is arranged in a rod-shaped helical assembly, giving rise to an oligomeric RNase ‘active’ complex. More recently, a yeast Ire1–quercetin–ADP complex (Wiseman et al, 2010) showed essentially the same back-to-back orientation that has been seen with the previous structures. All of the yeast structures have been crystallised from similar constructs that have an internal ∼24 amino-acid deletion within the kinase domain. Furthermore, all the structures are derived from phosphorylated yeast Ire1 protein and all describe the RNase ‘active’ conformation of yeast Ire1. We have now determined the crystal structure of the cytoplasmic kinase-RNase region of human Ire1α at 2.7 Å resolution, in a dephosphorylated state, and in a complex with Mg2+-ADP, revealing a face-to-face dimer of the kinase domains, reminiscent of dimerisation-activated kinases such as Chk2, LOK, SLK, and DAPK3 (Oliver et al, 2006, 2007; Pike et al, 2008), in which the kinase active site and activation segment of the Ire1 protomers are in a suitable orientation and proximity for transphosphorylation. We believe that this orientation is an early state ‘phosphoryl-transfer’ competent conformation of Ire1, which precedes and is distinct from the RNase ‘active’ back-to-back conformation, that has been visualised in the yeast structures. Furthermore, our biochemical analysis of human Ire1α shows direct dependence of RNase activity on autophosphorylation, and that inhibition of autophosphorylation by ATP-competitors blocks Ire1α RNase activity in vivo. Our combined structural and biochemical data show that human Ire1 behaves mechanistically like a typical receptor protein kinase and that inhibition of its kinase activity is likely to provide an important new therapeutic approach to multiple myeloma. Results Crystal structure of human Ire1α A cytoplasmic region of human Ire1α (residues 547–977) was expressed in insect cells, chromatographically purified, dephosphorylated and crystallised by vapour diffusion in a buffer containing Mg2+-ADP. Diffraction data to 2.7 Å were collected on station I03 at the Diamond Light Source and the structure was phased by molecular replacement using the structure of yeast Ire1. The crystals contain four molecules of Ire1α in the asymmetric unit (see Materials and methods). The human Ire1α cytoplasmic region has a very similar overall structure to those previously determined for yeast Ire1 (Lee et al, 2008; Korennykh et al, 2009; Wiseman et al, 2010) (Figure 1A), consisting of a bilobal protein kinase fold with the ATP-binding site in the cleft between the N-terminal β-sheet lobe and the α-helical C-terminal lobe. A distinct feature of Ire1 is an addition helical domain, fused to and structurally continuous with the C-terminal lobe of the kinase, which provides the sequence-specific ribonuclease activity by which Ire1 is able to initiate splicing of Xbp1 mRNA (Sidrauski and Walter, 1997; Tirasophon et al, 2000). Figure 1.Structure of human Ire1α. (A) Secondary structure cartoons of the crystal structures of cytoplasmic regions of human Ire1α (left) and yeast Ire1 (middle and right), rainbow coloured blue → red, N → C-terminus. Non-canonical secondary structural elements in the yeast structures (αD′, αE′), which mediate protein–protein interactions in the different yeast Ire1 crystal lattices, are not conserved in human Ire1α. Both yeast crystal structures have an engineered loop (Δ24 loop) connecting α-helices EF and F, from which 24 residues were deleted to obtain crystals. The equivalent loop in human (and other metazoan) Ire1α is much shorter, and was not modified. A small helix in the C-terminal RNase domain (α3′), which has been suggested to contribute to the catalytic function of this domain is fully ordered in human Ire1α. Human Ire1α contains an additional β-strand (βC′) formed by the unwinding of the N-terminal end of the kinase domain C-helix, which makes a β-sheet interaction with the N-terminal end of the activation segment. (B) Close-up of the β-sheet interaction between the βC′-strand and the N-terminal part of the activation segment immediately following the conserved DFG motif at residues 711–713. This β-sheet interaction directs the subsequent residues of the activation segment away from the body of the kinase, towards the kinase active site of a second Ire1α molecule (see below). To our knowledge, this type of interaction, involving the unwinding of the C-helix, has not previously been described in protein kinases. The tip of the activation segment between residues 720 and 730 is poorly ordered. (C) Close-up of Mg2+-ADP bound to dephosphorylated human Ire1α. The DFG motif is in the ‘in’ conformation associated with an activated kinase, and the side chain of Asp711 in the motif provides a direct ligand interaction with the Mg2+ ion. Electron density mesh is from a Fo–Fc difference Fourier map calculated with Mg-ADP omitted, and contoured at 3.0 σ. Download figure Download PowerPoint The human Ire1α structure is visible from residue 561 to 966, which is 11 residues upstream of the biological C-terminus. The central ∼10 residues (720–729) of the kinase activation segment, containing a putative phosphorylation site at Ser724, conserved in yeast Ire1, are disordered. The activation segment is similarly disordered in the high-resolution yeast structure (2RIO) and Ire1–quercetin complex structure (3LJ0), but is ordered in the low-resolution oligomeric yeast structure (3FBV), where it projects sideways and forms an interface with another monomer. The conformation and orientation of the yeast activation segments are markedly different to human Ire1α (see below). Residues 746–748 in the human Ire1α kinase domain, at the tip of the loop connecting helix αF and the short αEF helix carrying the APE motif, are also disordered. In yeast Ire1, this loop is substantially longer, and was artificially ‘short circuited’ by excision of 24 residues in the published structures of the yeast enzyme, which contributes to crystal lattice interactions. The connection between helices 3 and 4 in the RNase domain, which was disordered in the higher-resolution yeast Ire1 structure, is fully ordered in human Ire1α where it has a helical conformation. Two helical segments (αD′ and αE′) observed in the yeast Ire1 structures (Figure 1A) do not occur in human Ire1, where the equivalent segments are short loops. The most significant differences between the human Ire1 and yeast Ire1 monomer structures occur in the N-lobe and cleft region of the kinase domain. The kinase N-lobe overall is rotated ∼6° in human Ire1 with respect to the C-lobe, compared with the juxtaposition of these lobes in the yeast structures. Accompanying this is a substantial unravelling of the N-terminal end of the C-helix in the Ire1 kinase N-lobe, which adopts a fully ordered extended conformation for residues 602–611. The rest of this segment, which maintains a helical conformation, is displaced from its position in yeast Ire1 by ∼6 Å, separating the conserved Lys-Glu ion pair, required for activity (Huse and Kuriyan, 2002). The DFG motif, however, has an ‘in’ conformation, consistent with a catalytically competent state of the protein, with the carboxylate side chain of Asp711 interacting with the Mg2+ associated with the bound ADP (Figure 1B and C). In both the human and the two yeast Ire1 structures (3LJ0, 2RIO), as in many protein kinases in their unphosphorylated states, the activation segment (Johnson et al, 1996) connecting the protein kinase DFG and APE motifs is disordered. However, in the lower-resolution phosphorylated yeast Ire1, an ordered activation segment projects out and away from the active site and forms an interface with another monomer in a ‘side-to-side’ orientation. The phosphorylated residues are intimately involved in this interaction, with the invariant pSer841 interacting with Lys678 from the opposing monomer, and pSer840 associating with Asp763 and Glu764 in a negatively charged pocket formed by the αD–αE loop. This arrangement is not seen in the human structure, since the orientation of the molecules relative to each other is different to that seen in all the yeast structures (see below). In the higher-resolution yeast Ire1 structure, the ordered N-terminal part of the activation segment, extending from the DFG motif, is directed away from the bound ADP, in the same direction as the Δ24-αEF–αF loop. The C-terminal part leading into the APE motif and the Δ24–αEF–αF loop heads in the opposite direction, so that the ends of the disordered 15 residues of the activation segment, that connects them, are separated by >30 Å. By contrast, in human Ire1α, the N-terminal end of the activation segment immediately following the DFG motif projects directly out from the body of the kinase towards the active site of the opposing monomer, making an anti-parallel β-sheet with the residues from the N-terminal end of the C-helix, in their unwound extended conformation. The C-terminal segment upstream of the APE motif heads in the same direction, so that the visible tips of the ordered parts of the human activation segment are only ∼9 Å apart, and could be readily bridged by the intervening residues making a reverse loop (Figure 1B). Ire1 dimerisation and autophosphorylation In the current models for the activation mechanism of Ire1, the protein is believed to be held in a monomeric and inactive state by the ER Hsp70 homologue Bip, which binds to the ER luminal domain of Ire1 (Bertolotti et al, 2000; Okamura et al, 2000; Kimata et al, 2003). Accumulation of unfolded protein in the ER promotes release of Bip, permitting dimerisation of the luminal domains (Credle et al, 2005; Zhou et al, 2006). This in turn drives oligomeric association and transphosphorylation of the cytoplasmic kinase domain (Shamu and Walter, 1996; Welihinda and Kaufman, 1996), which is an essential prerequisite for activation of the splicing activity of the cytoplasmic ribonuclease domain (Tirasophon et al, 2000). The requirement for formation of dimers (and possibly higher oligomers) in the activation of Ire1 has dominated the mechanistic interpretation of the previously described crystal structures of yeast Ire1 (Lee et al, 2008; Korennykh et al, 2009). In both studies (which utilised extremely similar constructs incorporating the engineered Δ24–αEF–αF loop), an intimate dimer could be identified in the crystal lattice, in which two Ire1 molecules formed a back-to-back interaction, with the nucleotide-binding site facing outwards. In both cases, this dimeric arrangement is further expanded by helical symmetry to generate the full crystal lattice—in the high-resolution structure (Lee et al, 2008), the helix has a six-fold periodicity and is consequently subsumed in the space group description of the lattice, while in the lower-resolution structure (Korennykh et al, 2009), the helix has a much longer periodicity and 14 copies of the back-to-back dimer constitute the crystallographic asymmetric unit. The arrangement of molecules in the human Ire1α crystals is quite different to that of the yeast Ire1 structures. Only two protein–protein interfaces of any substance are observed. The largest (burying >1700 Å2 of molecular surface) is an approximately two-fold symmetric parallel interaction, involving residues from the kinase domain only, and generates a ‘face-to-face’ dimer arrangement in which the open mouths of the interlobe clefts, containing bound Mg2+-ADP, face each other across the dimer (Figure 2A). The next largest interface (burying ∼1600 Å2) is an anti-parallel arrangement, involving the last part of the kinase C-lobe and the ribonuclease domain. This arrangement would be incompatible with the membrane tethering of Ire1 and is likely to be a crystal lattice contact only. Figure 2.Face-to-face kinase domain dimerisation of human Ire1α. (A) Human Ire1α forms a dimer in which the kinase active sites of the two monomers face each other. The activation segment of one monomer is directed towards the other, so that the target substrate residue, Ser724 would come into close proximity of the Mg2+-ATP bound in the opposite active site, and be phosphorylated by it. This arrangement of Ire1α molecules provides a straightforward mechanistic model for how dimerisation of Ire1 N-terminal domains in the lumen of the ER would facilitate association and transphosphorylation of their associated kinase domains on the cytoplasmic side of the membrane. (B) Autophosphorylation of dephosphorylated dimer interface mutants Q636A and F637A Ire1 as compared with wild type. Both wild-type and mutant proteins were incubated with 5 mM MgCl and 5 mM ATP at 37°C and samples were run at specific time points. Protein samples were visualised by western blot with generic Ire1α or the phospho-specific pS724-Ire1α. Download figure Download PowerPoint No interaction resembling the back-to-back dimer seen in the yeast structures can be identified in the human Ire1α crystal structure. This maybe in part due to features such as the αE′ helix, which has a major role in forming the yeast back-to-back dimer but is absent from human Ire1α (and all metazoan homologues), and only some of the residues involved in the back-to-back interaction of the yeast ribonuclease domain are conserved in the human enzyme. The additional side-to-side interface observed in the other yeast structure (3FBV), and invoked in a formation of a biologically important high-order Ire1 assembly, also depends upon interactions of a secondary structural element (helix αD′) that is absent from human (and other metazoan) Ire1. Whether the back-to-back orientation exists in the human Ire1 mechanism remains to be seen. However, what is certain is that the face-to-face orientation is distinct from the back-to-back arrangement visualised in the yeast structures. As the human protein was extensively dephosphorylated prior to crystallisation, this face-to-face arrangement most likely represents a functional state that precedes the back-to-back ‘RNase active’ arrangement seen in the phosphorylated yeast Ire1 structures. The face-to-face interaction observed in the human Ire1α crystals provides a straightforward mechanistic explanation for how the cytoplasmic kinase domains of two Ire1 molecules, brought into close proximity by the association of their luminal domains on the opposite side of the ER membrane, might achieve transphosphorylation. The conformation of the activation segment, bolstered and supported by the partial unwinding of the C-helix, directs the conserved phosphorylation target Ser724 at the tip of the activation segment from one molecule, towards the active site of the opposite Ire1α molecule in the face-to-face dimer (Figure 2A). This type of reciprocal face-to-face transphosphorylation has been demonstrated structurally and biochemically for a number of kinases from different branches of the kinome (e.g. Chk2, LOK, SLK, DAPK3, etc), which like Ire1 are activated by dimerisation (Oliver et al, 2006, 2007; Pike et al, 2008). The secondary structural elements involved in the human Ire1α face-to-face interaction (the β1–β2 hairpin, β3–αC loop, and αG helix) are present in all Ire1 homologues, and many of the interfacial residues are invariant or only conservatively varied in metazoa and yeast, suggesting that, this face-to-face interaction would be available to all Ire1 homologues. To test whether this interface is indeed biologically significant in the autophosphorylation process, we made mutations in key interfacial residues, Q636A and F637A, and determined their ability to autophosphorylate as compared with wild type (Figure 2B). Ire1 autophosphorylation was measured over several time points and using an antibody developed to measure phosphorylation of Ser724 on the activation segment of human Ire1α (see below). Both wild-type and mutant protein were extensively dephosphorylated prior to incubation with ATP-Mg2. Mutation of residues implicated in the face-to-face interaction observed in the Ire1α crystals, significantly diminishes the ability of Ire1 to autophosphorylate compared with wild type, indicating the importance of this interface. In particular, the mutant Q636A, which disrupts the hydrogen bond with Asp634 and its Gln637 counterpart in the other monomer, shows significant retardation of autophosphorylation especially over the initial 30 min compared with wild type. The mutation of phenylalanine to a lesser hydrophobic alanine has less of an impact, but is still significantly slower to autophosphorylate than wild type. Ire1α nucleotide binding Previous structural and biochemical studies have suggested a mechanisms of Ire1 activation, whereby ADP binding to the phosphorylated protein promotes a conform
Referência(s)