Structure (v.13, #5)

Kay has developed a new model for the structure of the transmembrane helices of an α/β integrin in the resting conformation and suggests how the interactions between these helices help transmit signals across the membrane.

Allosteric Control of O2 Reactivity in Rieske Oxygenases by John D. Lipscomb; Brian M. Hoffman (684-685).
Oxygen is Nature’s perfect reagent. On one hand, it is potentially a very strong oxidant. On the other hand, this potential is caged because the two highest energy valence electrons of the O2 molecule are unpaired. As a result, O2 is relatively unreactive with most other molecules, as almost all of these have paired electrons. Consequently, by modulating the properties of the O2 valence electrons, Nature can generate a reactive species under controlled conditions, catalyzing difficult reactions while still rigorously enforcing specificity. Special sets of enzymes termed oxygenases and oxidases have evolved to perform this task.

In this issue of Structure, show that Cox17, a mitochondrial chaperone supplying copper ions for the biogenesis of cytochrome c oxidase, exists in multiple, redox-dependent isomeric forms in the intermembrane space.

Circular Proteins: Ring around with NOESY by Hans J. Vogel; David I. Chan (688-690).
The secrets of ribosomally synthesized circular proteins are slowly revealed by gene sequencing and solution NMR studies of novel cyclotides and their precursors, as demonstrated by in this issue of Structure.

Processing of a 22 kDa Precursor Protein to Produce the Circular Protein Tricyclon A by Jason P. Mulvenna; Lillian Sando; David J. Craik (691-701).
Cyclotides are a family of plant proteins that have the unusual combination of head-to-tail backbone cyclization and a cystine knot motif. They are exceptionally stable and show resistance to most chemical, physical, and enzymatic treatments. The structure of tricyclon A, a previously unreported cyclotide, is described here. In this structure, a loop that is disordered in other cyclotides forms a β sheet that protrudes from the globular core. This study indicates that the cyclotide fold is amenable to the introduction of a range of structural elements without affecting the cystine knot core of the protein, which is essential for the stability of the cyclotides. Tricyclon A does not possess a hydrophobic patch, typical of other cyclotides, and has minimal hemolytic activity, making it suitable for pharmaceutical applications. The 22 kDa precursor protein of tricyclon A was identified and provides clues to the processing of these fascinating miniproteins.

One of the hallmark features of the integrin receptors is the ability to transmit signals bidirectionally through the cell membrane. The transmembrane integrin domains are pivotal to the signaling events. An understanding of the signaling mechanism requires structural information. Here, we report a structural model of the transmembrane and part of the cytosolic domains of the αIIbβ3 integrin in its resting state. The model was obtained computationally by a restrained conformational search of helix-helix interactions. It agrees with one published NMR structure of the cytoplasmic complex and can put many experimental findings on structural grounds. According to our model, integrins form an intricately designed coiled-coil structure in the resting state. The conserved Glycophorin A (GpA)-like sequence motif of the α, but not the β, subunit, is in the interface of this model. Based on our calculations and other data, a signaling mechanism that involves a transient GpA-like structure is proposed.

Folding Studies of Cox17 Reveal an Important Interplay of Cysteine Oxidation and Copper Binding by Fabio Arnesano; Erica Balatri; Lucia Banci; Ivano Bertini; Dennis R. Winge (713-722).
Cox17 is a key mitochondrial copper chaperone involved in the assembly of cytochrome c oxidase (COX). The NMR solution structure of the oxidized apoCox17 isoform consists of a coiled-coil conformation stabilized by two disulfide bonds involving Cys26/Cys57 and Cys36/Cys47. This appears to be a conserved tertiary fold of a class of proteins, localized within the mitochondrial intermembrane space, that contain a twin Cys-x9-Cys sequence motif. An isomerization of one disulfide bond from Cys26/Cys57 to Cys24/Cys57 is required prior to Cu(I) binding to form the Cu1Cox17 complex. Upon further oxidation of the apo-protein, a form with three disulfide bonds is obtained. The reduction of all disulfide bonds provides a molten globule form that can convert to an additional conformer capable of binding up to four Cu(I) ions in a polycopper cluster. This form of the protein is oligomeric. These properties are framed within a complete model of mitochondrial import and COX assembly.

SOMO (SOlution MOdeler) by Nithin Rai; Marcelo Nöllmann; Bruno Spotorno; Giovanni Tassara; Olwyn Byron; Mattia Rocco (723-734).
Reduced numbers of frictional/scattering centers are essential for tractable hydrodynamic and small-angle scattering data modeling. We present a method for generating medium-resolution models from the atomic coordinates of proteins, basically by using two nonoverlapping spheres of differing radii per residue. The computed rigid-body hydrodynamic parameters of BPTI, RNase A, and lysozyme models were compared with a large database of critically assessed experimental values. Overall, very good results were obtained, but significant discrepancies between X-ray- and NMR-derived models were found. Interestingly, they could be accounted for by properly considering the extent to which highly mobile surface side chains differently affect translational/rotational properties. Models of larger structures, such as fibrinogen fragment D and citrate synthase, also produced consistent results. Foremost among this method’s potential applications is the overall conformation and dynamics of modular/multidomain proteins and of supramolecular complexes. The possibility of merging data from high- and low-resolution structures greatly expands its scope.

The cyanobacterial circadian oscillator consists of three Kai proteins, KaiA, KaiB, and KaiC, in its oscillation feedback loop. Structural comparison reveals that the Kai system resembles the F1-ATPase system in which KaiC is equivalent to α3β3, KaiA to γδϵ, and KaiB to its inhibitory factor. It also suggests that there exists a possible haemagglutinin-like spring-loaded mechanism for the activation of KaiA during the formation of Kai complexes.

The AXH Domain Adopts Alternative Folds by Cesira de Chiara; Rajesh P. Menon; Salvatore Adinolfi; Jasper de Boer; Eleni Ktistaki; Geoff Kelly; Lesley Calder; Dimitris Kioussis; Annalisa Pastore (743-753).
AXH is a protein module identified in two unrelated families that comprise the transcriptional repressor HBP1 and ataxin-1 (ATX1), the protein responsible for spinocerebellar ataxia type-1 (SCA1). SCA1 is a neurodegenerative disorder associated with protein misfolding and formation of toxic intranuclear aggregates. We have solved the structure in solution of monomeric AXH from HBP1. The domain adopts a nonclassical permutation of an OB fold and binds nucleic acids, a function previously unidentified for this region of HBP1. Comparison of HBP1 AXH with the crystal structure of dimeric ATX1 AXH indicates that, despite the significant sequence homology, the two proteins have different topologies, suggesting that AXH has chameleon properties. We further demonstrate that HBP1 AXH remains monomeric, whereas the ATX1 dimer spontaneously aggregates and forms fibers. Our results describe an entirely novel, to our knowledge, example of a chameleon fold and suggest a link between these properties and the SCA1 pathogenesis.

Scorpion-Toxin Mimics of CD4 in Complex with Human Immunodeficiency Virus gp120 by Chih-chin Huang; François Stricher; Loïc Martin; Julie M. Decker; Shahzad Majeed; Philippe Barthe; Wayne A. Hendrickson; James Robinson; Christian Roumestand; Joseph Sodroski; Richard Wyatt; George M. Shaw; Claudio Vita; Peter D. Kwong (755-768).
The binding surface on CD4 for the HIV-1 gp120 envelope glycoprotein has been transplanted previously onto a scorpion-toxin scaffold. Here, we use X-ray crystallography to characterize atomic-level details of gp120 with this transplant, CD4M33. Despite known envelope flexibility, the conformation of gp120 induced by CD4M33 was so similar to that induced by CD4 that localized measures were required to distinguish ligand-induced differences from lattice variation. To investigate relationships between structure, function, and mimicry, an F23 analog of CD4M33 was devised. Structural and thermodynamic analyses showed F23 to be a better molecular mimic of CD4 than CD4M33. F23 also showed increased neutralization breadth, against diverse isolates of HIV-1, HIV-2, and SIVcpz. Our results lend insight into the stability of the CD4 bound conformation of gp120, define measures that quantify molecular mimicry as a function of evolutionary distance, and suggest how such evaluations might be useful in developing mimetic antagonists with increased neutralization breadth.

The Active Conformation of the PAK1 Kinase Domain by Ming Lei; Michael A. Robinson; Stephen C. Harrison (769-778).
The p21-activated kinases (PAKs) participate in cytoskeletal control networks, downstream of Rho-family GTPases. A structure of PAK1 in an autoregulated, “off” state showed that a regulatory region, N-terminal to the kinase domain, forces the latter into an inactive conformation, prevents phosphorylation of Thr423 in the activation loop, and promotes dimerization. We have now determined structures at 1.8 Å resolution for the free PAK1 kinase domain, with a mutation in the active site that blocks enzymatic activity, and for the same domain with a “phosphomimetic” mutation in the activation loop. The two very similar structures show that even in the absence of a phosphorylated Thr423, the kinase has an essentially active conformation. When Cdc42 binds the regulatory region and dissociates the dimer, PAK1 will be in an “intermediate-active” state, with a capacity to phosphorylate itself or other substrates even prior to modification of its activation loop.

Crystal Structure of a RuBisCO-like Protein from the Green Sulfur Bacterium Chlorobium tepidum by Huiying Li; Michael R. Sawaya; F. Robert Tabita; David Eisenberg (779-789).
Ribulose 1,5-bisphosphate carboxylase/oxygenase (RuBisCO) catalyzes the incorporation of atmospheric CO2 into ribulose 1,5-bisphosphate (RuBP). RuBisCOs are classified into four forms based on sequence similarity: forms I, II and III are bona fide RuBisCOs; form IV, also called the RuBisCO-like protein (RLP), lacks several of the substrate binding and catalytic residues and does not catalyze RuBP-dependent CO2 fixation in vitro. To contribute to understanding the function of RLPs, we determined the crystal structure of the RLP from Chlorobium tepidum. The overall structure of the RLP is similar to the structures of the three other forms of RuBisCO; however, the active site is distinct from those of bona fide RuBisCOs and suggests that the RLP is possibly capable of catalyzing enolization but not carboxylation. Bioinformatic analysis of the protein functional linkages suggests that this RLP coevolved with enzymes of the bacteriochlorophyll biosynthesis pathway and may be involved in processes related to photosynthesis.

Implications for Switching Restriction Enzyme Specificities from the Structure of BstYI Bound to a BglII DNA Sequence by Sharon A. Townson; James C. Samuelson; Shuang-yong Xu; Aneel K. Aggarwal (791-801).
The type II restriction endonuclease BstYI recognizes the degenerate sequence 5′-RGATCY-3′ (where R = A/G and Y = C/T), which overlaps with both BamHI (GGATCC) and BglII (AGATCT), and thus raises the question of whether BstYI DNA recognition will be more BamHI-like or BglII-like. We present here the structure of BstYI bound to a cognate DNA sequence (AGATCT). We find the complex to be more BglII-like with similarities mapping to DNA conformation, domain organization, and residues involved in catalysis. However, BstYI is unique in containing an extended arm subdomain, and the mechanism of DNA capture has both BglII-like and BamHI-like elements. Further, DNA recognition is more minimal than BglII and BamHI, where only two residues mediate recognition of the entire core sequence. Taken together, the structure reveals a mechanism of degenerate DNA recognition and offers insights into the possibilities and limitations in altering specificities of closely related restriction enzymes.

Structural Studies of the Parainfluenza Virus 5 Hemagglutinin-Neuraminidase Tetramer in Complex with Its Receptor, Sialyllactose by Ping Yuan; Thomas B. Thompson; Beth A. Wurzburg; Reay G. Paterson; Robert A. Lamb; Theodore S. Jardetzky (803-815).
The paramyxovirus hemagglutinin-neuraminidase (HN) functions in virus attachment to cells, cleavage of sialic acid from oligosaccharides, and stimulating membrane fusion during virus entry into cells. The structural basis for these diverse functions remains to be fully understood. We report the crystal structures of the parainfluenza virus 5 (SV5) HN and its complexes with sialic acid, the inhibitor DANA, and the receptor sialyllactose. SV5 HN shares common structural features with HN of Newcastle disease virus (NDV) and human parainfluenza 3 (HPIV3), but unlike the previously determined HN structures, the SV5 HN forms a tetramer in solution, which is thought to be the physiological oligomer. The sialyllactose complex reveals intact receptor within the active site, but no major conformational changes in the protein. The SV5 HN structures do not support previously proposed models for HN action in membrane fusion and suggest alternative mechanisms by which HN may promote virus entry into cells.

2-Oxoquinoline 8-monooxygenase is a Rieske non-heme iron oxygenase that catalyzes the NADH-dependent oxidation of the N-heterocyclic aromatic compound 2-oxoquinoline to 8-hydroxy-2-oxoquinoline in the soil bacterium Pseudomonas putida 86. The crystal structure of the oxygenase component of 2-oxoquinoline 8-monooxygenase shows a ring-shaped, C3-symmetric arrangement in which the mononuclear Fe(II) ion active site of one monomer is at a distance of 13 Å from the Rieske-[2Fe-2S] center of a second monomer. Structural analyses of oxidized, reduced, and substrate bound states reveal the molecular bases for a new function of Fe-S clusters. Reduction of the Rieske center modulates the mononuclear Fe through a chain of conformational changes across the subunit interface, resulting in the displacement of Fe and its histidine ligand away from the substrate binding site. This creates an additional coordination site at the mononuclear Fe(II) ion and can open a pathway for dioxygen to bind in the substrate-containing active site.

The angiopoietins comprise a small class of secreted glycoproteins that play crucial roles in the maturation and maintenance of the mammalian vascular and lymphatic systems. They exert their effects through a member of the tyrosine kinase receptor family, Tie2. Angiopoietin/Tie2 signaling is unique among tyrosine kinase receptor-ligand systems in that distinct angiopoietin ligands, although highly homologous, can function as agonists or antagonists in a context-dependent manner. In an effort to understand this molecular dichotomy, we have crystallized and determined the 2.4 Å crystal structure of the Angiopoietin-2 (Ang2) receptor binding region. The structure reveals a fibrinogen fold with a unique C-terminal P domain. Conservation analysis and structure-based mutagenesis identify a groove on the Ang2 molecular surface that mediates receptor recognition.