Structural basis for nuclear import complex dissociation by RanGTP

Nuclear protein import is mediated mainly by the transport factor importin-beta that binds cytoplasmic cargo, most often via the importin-alpha adaptor, and then transports it through nuclear pore complexes. This active transport is driven by disassembly of the import complex by nuclear RanGTP. The switch I and II loops of Ran change conformation with nucleotide state, and regulate its interactions with nuclear trafficking components. Importin-beta consists of 19 HEAT repeats that are based on a pair of antiparallel alpha-helices (referred to as the A- and B-helices). The HEAT repeats stack to yield two C-shaped arches, linked together to form a helicoidal molecule that has considerable conformational flexibility. Here we present the structure of full-length yeast importin-beta (Kap95p or karyopherin-beta) complexed with RanGTP, which provides a basis for understanding the crucial cargo-release step of nuclear import. We identify a key interaction site where the RanGTP switch I loop binds to the carboxy-terminal arch of Kap95p. This interaction produces a change in helicoidal pitch that locks Kap95p in a conformation that cannot bind importin-alpha or cargo. We suggest an allosteric mechanism for nuclear import complex disassembly by RanGTP.     

Structure and assembly of the Alu domain of the mammalian signal recognition particle

The Alu domain of the mammalian signal recognition particle (SRP) comprises the heterodimer of proteins SRP9 and SRP14 bound to the 5' and 3' terminal sequences of SRP RNA. It retards the ribosomal elongation of signal-peptide-containing proteins before their engagement with the translocation machinery in the endoplasmic reticulum. Here we report two crystal structures of the heterodimer SRP9/14 bound either to the 5' domain or to a construct containing both 5' and 3' domains. We present a model of the complete Alu domain that is consistent with extensive biochemical data. SRP9/14 binds strongly to the conserved core of the 5' domain, which forms a U-turn connecting two helical stacks. Reversible docking of the more weakly bound 3' domain might be functionally important in the mechanism of translational regulation. The Alu domain structure is probably conserved in other cytoplasmic ribonucleoprotein particles and retroposition intermediates containing SRP9/14-bound RNAs transcribed from Alu repeats or related elements in genomic DNA.     

Structure of the nuclear transport complex karyopherin-beta2-Ran x GppNHp.

Transport factors in the karyopherin-beta (also called importin-beta) family mediate the movement of macromolecules in nuclear-cytoplasmic transport pathways. Karyopherin-beta2 (transportin) binds a cognate import substrate and targets it to the nuclear pore complex. In the nucleus, Ran x GTP binds karyopherin-beta2 and dissociates the substrate. Here we present the 3.0 A structure of the karyopherin-beta2-Ran x GppNHp complex where GppNHp is a non-hydrolysable GTP analogue. Karyopherin-beta2 contains eighteen HEAT repeats arranged into two continuous orthogonal arches. Ran is clamped in the amino-terminal arch and substrate-binding activity is mapped to the carboxy-terminal arch. A large loop in HEAT repeat 7 spans both arches. Interactions of the loop with Ran and the C-terminal arch implicate it in GTPase-mediated dissociation of the import-substrate. Ran x GppNHp in the complex shows extensive structural rearrangement, compared to Ran GDP, in regions contacting karyopherin-beta2. This provides a structural basis for the specificity of the karyopherin-beta family for the GTP-bound state of Ran, as well as a rationale for interactions of the karyopherin-Ran complex with the regulatory proteins ranGAP, ranGEF and ranBP1.     

Structural basis for the assembly of a nuclear export complex

The nuclear import and export of macromolecular cargoes through nuclear pore complexes is mediated primarily by carriers such as importin-beta. Importins carry cargoes into the nucleus, whereas exportins carry cargoes to the cytoplasm. Transport is orchestrated by nuclear RanGTP, which dissociates cargoes from importins, but conversely is required for cargo binding to exportins. Here we present the 2.0 A crystal structure of the nuclear export complex formed by exportin Cse1p complexed with its cargo (Kap60p) and RanGTP, thereby providing a structural framework for understanding nuclear protein export and the different functions of RanGTP in export and import. In the complex, Cse1p coils around both RanGTP and Kap60p, stabilizing the RanGTP-state and clamping the Kap60p importin-beta-binding domain, ensuring that only cargo-free Kap60p is exported. Mutagenesis indicated that conformational changes in exportins couple cargo binding to high affinity for RanGTP, generating a spring-loaded molecule to facilitate disassembly of the export complex following GTP hydrolysis in the cytoplasm.     

Structure of the pancreatic lipase-procolipase complex.

Interfacial adsorption of pancreatic lipase is strongly dependent on the physical chemical properties of the lipid surface. These properties are affected by amphiphiles such as phospholipids and bile salts. In the presence of such amphiphiles, lipase binding to the interface requires a protein cofactor, colipase. We obtained crystals of the pancreatic lipase-procolipase complex and solved the structure at 3.04 A resolution. Here we describe the structure of procolipase, which essentially consists of three 'fingers' and is topologically comparable to snake toxins. The tips of the fingers contain most of the hydrophobic amino acids and presumably form the interfacial binding site. Lipase binding occurs at the opposite side to this site and involves polar interactions. Determination of the three-dimensional structure of pancreatic lipase has revealed the presence of two domains: an amino-terminal domain, at residues 1-336 containing the active site and a carboxy-terminal domain at residues 337-449 (ref. 6). Procolipase binds exclusively to the C-terminal domain of lipase. No conformational change in the lipase molecule is induced by the binding of procolipase.     

Crystal structure of nerve growth factor in complex with the ligand-binding domain of the TrkA receptor.

Nerve growth factor (NGF) is involved in a variety of processes involving signalling, such as cell differentiation and survival, growth cessation and apoptosis of neurons. These events are mediated by NGF as a result of binding to its two cell-surface receptors, TrkA and p75. TrkA is a receptor with tyrosine kinase activity that forms a high-affinity binding site for NGF. Of the five domains comprising its extracellular portion, the immunoglobulin-like domain proximal to the membrane (TrkA-d5 domain) is necessary and sufficient for NGF binding. Here we present the crystal structure of human NGF in complex with human TrkA-d5 at 2.2 A resolution. The ligand-receptor interface consists of two patches of similar size. One patch involves the central beta-sheet that forms the core of the homodimeric NGF molecule and the loops at the carboxy-terminal pole of TrkA-d5. The second patch comprises the amino-terminal residues of NGF, which adopt a helical conformation upon complex formation, packing against the 'ABED' sheet of TrkA-d5. The structure is consistent with results from mutagenesis experiments for all neurotrophins, and indicates that the first patch may constitute a conserved binding motif for all family members, whereas the second patch is specific for the interaction between NGF and TrkA.     

Structure of the Sec23/24-Sar1 pre-budding complex of the COPII vesicle coat

COPII-coated vesicles form on the endoplasmic reticulum by the stepwise recruitment of three cytosolic components: Sar1-GTP to initiate coat formation, Sec23/24 heterodimer to select SNARE and cargo molecules, and Sec13/31 to induce coat polymerization and membrane deformation. Crystallographic analysis of the Saccharomyces cerevisiae Sec23/24-Sar1 complex reveals a bow-tie-shaped structure, 15 nm long, with a membrane-proximal surface that is concave and positively charged to conform to the size and acidic-phospholipid composition of the COPII vesicle. Sec23 and Sar1 form a continuous surface stabilized by a non-hydrolysable GTP analogue, and Sar1 has rearranged from the GDP conformation to expose amino-terminal residues that will probably embed in the bilayer. The GTPase-activating protein (GAP) activity of Sec23 involves an arginine side chain inserted into the Sar1 active site. These observations establish the structural basis for GTP-dependent recruitment of a vesicular coat complex, and for uncoating through coat-controlled GTP hydrolysis.     

Crystal structure of the FimD usher bound to its cognate FimC-FimH substrate

Type 1 pili are the archetypal representative of a widespread class of adhesive multisubunit fibres in Gram-negative bacteria. During pilus assembly, subunits dock as chaperone-bound complexes to an usher, which catalyses their polymerization and mediates pilus translocation across the outer membrane. Here we report the crystal structure of the full-length FimD usher bound to the FimC-FimH chaperone-adhesin complex and that of the unbound form of the FimD translocation domain. The FimD-FimC-FimH structure shows FimH inserted inside the FimD 24-stranded β-barrel translocation channel. FimC-FimH is held in place through interactions with the two carboxy-terminal periplasmic domains of FimD, a binding mode confirmed in solution by electron paramagnetic resonance spectroscopy. To accommodate FimH, the usher plug domain is displaced from the barrel lumen to the periplasm, concomitant with a marked conformational change in the β-barrel. The amino-terminal domain of FimD is observed in an ideal position to catalyse incorporation of a newly recruited chaperone-subunit complex. The FimD-FimC-FimH structure provides unique insights into the pilus subunit incorporation cycle, and captures the first view of a protein transporter in the act of secreting its cognate substrate.     

Subunit arrangement and phenylethanolamine binding in GluN1/GluN2B NMDA receptors

Since it was discovered that the anti-hypertensive agent ifenprodil has neuroprotective activity through its effects on NMDA (N-methyl-D-aspartate) receptors, a determined effort has been made to understand the mechanism of action and to develop improved therapeutic compounds on the basis of this knowledge. Neurotransmission mediated by NMDA receptors is essential for basic brain development and function. These receptors form heteromeric ion channels and become activated after concurrent binding of glycine and glutamate to the GluN1 and GluN2 subunits, respectively. A functional hallmark of NMDA receptors is that their ion-channel activity is allosterically regulated by binding of small compounds to the amino-terminal domain (ATD) in a subtype-specific manner. Ifenprodil and related phenylethanolamine compounds, which specifically inhibit GluN1 and GluN2B NMDA receptors, have been intensely studied for their potential use in the treatment of various neurological disorders and diseases, including depression, Alzheimer's disease and Parkinson's disease. Despite considerable enthusiasm, mechanisms underlying the recognition of phenylethanolamines and ATD-mediated allosteric inhibition remain limited owing to a lack of structural information. Here we report that the GluN1 and GluN2B ATDs form a heterodimer and that phenylethanolamine binds at the interface between GluN1 and GluN2B, rather than within the GluN2B cleft. The crystal structure of the heterodimer formed between the GluN1b ATD from Xenopus laevis and the GluN2B ATD from Rattus norvegicus shows a highly distinct pattern of subunit arrangement that is different from the arrangements observed in homodimeric non-NMDA receptors and reveals the molecular determinants for phenylethanolamine binding. Restriction of domain movement in the bi-lobed structure of the GluN2B ATD, by engineering of an inter-subunit disulphide bond, markedly decreases sensitivity to ifenprodil, indicating that conformational freedom in the GluN2B ATD is essential for ifenprodil-mediated allosteric inhibition of NMDA receptors. These findings pave the way for improving the design of subtype-specific compounds with therapeutic value for neurological disorders and diseases.     

Novel potential mitotic motor protein encoded by the fission yeast cut7+ gene

The structure equivalent to higher eukaryotic centrosomes in fission yeast, the nuclear membrane-bound spindle pole body, is inactive during interphase. On transition from G2 to M phase of the cell cycle, the spindle pole body duplicates; the daughter pole bodies seed microtubules which interdigitate to form a short spindle that elongates to span the nucleus at metaphase. We have identified two loci which, when mutated, block spindle formation. The predicted product of one of these genes, cut7+, contains an amino-terminal domain similar to the kinesin heavy chain head domain, indicating that the cut7+ product could be a spindle motor. The cut7+ gene resembles the Aspergillus nidulans putative spindle motor gene bimC, both in terms of its organization with a homologous amino-terminal head and no obvious heptad repeats and in the morphology of the mutant phenotype. But we find no similarity between the carboxy termini of these genes, suggested that either the cut7+ gene represents a new class of kinesin genes and that fission yeast may in addition contain a bimC homologue, or that the carboxy termini of these mitotic kinesins are not evolutionarily conserved and that the cut7+ gene belongs to a subgroup of bimC-related kinesins.     

The structures of HsIU and the ATP-dependent protease HsIU-HsIV

The degradation of cytoplasmic proteins is an ATP-dependent process. Substrates are targeted to a single soluble protease, the 26S proteasome, in eukaryotes and to a number of unrelated proteases in prokaryotes. A surprising link emerged with the discovery of the ATP-dependent protease HslVU (heat shock locus VU) in Escherichia coli. Its protease component HslV shares approximately 20% sequence similarity and a conserved fold with 20S proteasome beta-subunits. HslU is a member of the Hsp100 (Clp) family of ATPases. Here we report the crystal structures of free HslU and an 820,000 relative molecular mass complex of HslU and HslV-the first structure of a complete set of components of an ATP-dependent protease. HslV and HslU display sixfold symmetry, ruling out mechanisms of protease activation that require a symmetry mismatch between the two components. Instead, there is conformational flexibility and domain motion in HslU and a localized order-disorder transition in HslV. Individual subunits of HslU contain two globular domains in relative orientations that correlate with nucleotide bound and unbound states. They are surprisingly similar to their counterparts in N-ethylmaleimide-sensitive fusion protein, the prototype of an AAA-ATPase. A third, mostly alpha-helical domain in HslU mediates the contact with HslV and may be the structural equivalent of the amino-terminal domains in proteasomal AAA-ATPases.     

The crystal structure of DNA mismatch repair protein MutS binding to a G x T mismatch

DNA mismatch repair ensures genomic integrity on DNA replication. Recognition of a DNA mismatch by a dimeric MutS protein initiates a cascade of reactions and results in repair of the newly synthesized strand; however, details of the molecular mechanism remain controversial. Here we present the crystal structure at 2.2 A of MutS from Escherichia coli bound to a G x T mismatch. The two MutS monomers have different conformations and form a heterodimer at the structural level. Only one monomer recognizes the mismatch specifically and has ADP bound. Mismatch recognition occurs by extensive minor groove interactions causing unusual base pairing and kinking of the DNA. Nonspecific major groove DNA-binding domains from both monomers embrace the DNA in a clamp-like structure. The interleaved nucleotide-binding sites are located far from the DNA. Mutations in human MutS alpha (MSH2/MSH6) that lead to hereditary predisposition for cancer, such as hereditary non-polyposis colorectal cancer, can be mapped to this crystal structure.     

Structure of the Ku heterodimer bound to DNA and its implications for double-strand break repair

The Ku heterodimer (Ku70 and Ku80 subunits) contributes to genomic integrity through its ability to bind DNA double-strand breaks and facilitate repair by the non-homologous end-joining pathway. The crystal structure of the human Ku heterodimer was determined both alone and bound to a 55-nucleotide DNA element at 2.7 and 2.5 A resolution, respectively. Ku70 and Ku80 share a common topology and form a dyad-symmetrical molecule with a preformed ring that encircles duplex DNA. The binding site can cradle two full turns of DNA while encircling only the central 3-4 base pairs (bp). Ku makes no contacts with DNA bases and few with the sugar-phosphate backbone, but it fits sterically to major and minor groove contours so as to position the DNA helix in a defined path through the protein ring. These features seem well designed to structurally support broken DNA ends and to bring the DNA helix into phase across the junction during end processing and ligation.     

Structural basis for the regulation of tubulin by vinblastine

Vinblastine is one of several tubulin-targeting Vinca alkaloids that have been responsible for many chemotherapeutic successes since their introduction in the clinic as antitumour drugs. In contrast with the two other classes of small tubulin-binding molecules (Taxol and colchicine), the binding site of vinblastine is largely unknown and the molecular mechanism of this drug has remained elusive. Here we report the X-ray structure of vinblastine bound to tubulin in a complex with the RB3 protein stathmin-like domain (RB3-SLD). Vinblastine introduces a wedge at the interface of two tubulin molecules and thus interferes with tubulin assembly. Together with electron microscopical and biochemical data, the structure explains vinblastine-induced tubulin self-association into spiral aggregates at the expense of microtubule growth. It also shows that vinblastine and the amino-terminal part of RB3-SLD binding sites share a hydrophobic groove on the alpha-tubulin surface that is located at an intermolecular contact in microtubules. This is an attractive target for drugs designed to perturb microtubule dynamics by interfacial interference, for which tubulin seems ideally suited because of its propensity to self-associate.