E.Y. Jones

Crystal structure of a heparin- and integrin-binding segment of human fibronectin.

The crystal structure of human fibronectin (FN) type III repeats 12–14 reveals the primary heparin‐binding site, a clump of positively charged residues in FN13, and a putative minor site ∼60 Å away in FN14. The IDAPS motif implicated in integrin α4β1 binding is at the FN13–14 junction, rendering the critical Asp184 inaccessible to integrin. Asp184 clamps the BC loop of FN14, whose sequence (PRARI) is reminiscent of the synergy sequence (PHSRN) of FN9. Mutagenesis studies prompted by this observation reveal that both arginines of the PRARI sequence are important for α4β1 binding to FN12–14. The PRARI motif may represent a new class of integrin‐binding sites. The spatial organization of the binding sites suggests that heparin and integrin may bind in concert.

The ligand-binding face of the semaphorins revealed by the high-resolution crystal structure of SEMA4D

Semaphorins, proteins characterized by an extracellular sema domain, regulate axon guidance, immune function and angiogenesis. The crystal structure of SEMA4D (residues 1–657) shows the sema topology to be a seven-bladed β-propeller, revealing an unexpected homology with integrins. The sema β-propeller contains a distinctive 77-residue insertion between β-strands C and D of blade 5. Blade 7 is followed by a domain common to plexins, semaphorins and integrins (PSI domain), which forms a compact cysteine knot abutting the side of the propeller, and an Ig-like domain. The top face of the β-propeller presents prominent loops characteristic of semaphorins. In addition to limited contact between the Ig-like domains, the homodimer is stabilized through extensive interactions between the top faces in a sector of the β-propeller used for heterodimerization in integrins. This face of the propeller also mediates ligand binding in integrins, and functional data for semaphorin-receptor interactions map to the equivalent surface.

Structural basis of Nipah and Hendra virus attachment to their cell-surface receptor ephrin-B2

Nipah and Hendra viruses are emergent paramyxoviruses, causing disease characterized by rapid onset and high mortality rates, resulting in their classification as Biosafety Level 4 pathogens. Their attachment glycoproteins are essential for the recognition of the cell-surface receptors ephrin-B2 (EFNB2) and ephrin-B3 (EFNB3). Here we report crystal structures of both Nipah and Hendra attachment glycoproteins in complex with human EFNB2. In contrast to previously solved paramyxovirus attachment complexes, which are mediated by sialic acid interactions, the Nipah and Hendra complexes are maintained by an extensive protein-protein interface, including a crucial phenylalanine side chain on EFNB2 that fits snugly into a hydrophobic pocket on the viral protein. By analogy with the development of antivirals against sialic acid binding viruses, these results provide a structural template to target antiviral inhibition of protein-protein interactions.

Conflicting Selective Forces Affect T Cell Receptor Contacts in an Immunodominant Human Immunodeficiency Virus Epitope.

Cytotoxic T lymphocytes (CTLs) are critical for the control of human immunodeficiency virus, but containment of virus replication can be undermined by mutations in CTL epitopes that lead to virus escape. We analyzed the evolution in vivo of an immunodominant, HLA-A2–restricted CTL epitope and found two principal, diametrically opposed evolutionary pathways that exclusively affect T cell–receptor contact residues. One pathway was characterized by acquisition of CTL escape mutations and the other by selection for wild-type amino acids. The pattern of CTL responses to epitope variants shaped which variant(s) prevailed in the virus population. The pathways notably influenced the amount of plasma virus, as patients with efficient CTL selection had lower plasma viral loads than did patients without efficient selection. Thus, viral escape from CTL responses does not necessarily correlate with disease progression.

Molecular analysis of receptor protein tyrosine phosphatase μ‐mediated cell adhesion

Type IIB receptor protein tyrosine phosphatases (RPTPs) are bi‐functional cell surface molecules. Their ectodomains mediate stable, homophilic, cell‐adhesive interactions, whereas the intracellular catalytic regions can modulate the phosphorylation state of cadherin/catenin complexes. We describe a systematic investigation of the cell‐adhesive properties of the extracellular region of RPTPμ, a prototypical type IIB RPTP. The crystal structure of a construct comprising its N‐terminal MAM (meprin/A5/μ) and Ig domains was determined at 2.7 Å resolution; this assigns the MAM fold to the jelly‐roll family and reveals extensive interactions between the two domains, which form a rigid structural unit. Structure‐based site‐directed mutagenesis, serial domain deletions and cell‐adhesion assays allowed us to identify the four N‐terminal domains (MAM, Ig, fibronectin type III (FNIII)‐1 and FNIII‐2) as a minimal functional unit. Biophysical characterization revealed at least two independent types of homophilic interaction which, taken together, suggest that there is the potential for formation of a complex and possibly ordered array of receptor molecules at cell contact sites.

Structure of the Snake-Venom Toxin Convulxin

Snake venoms contain a number of proteins that interact with components of the haemostatic system that promote or inhibit events leading to blood-clot formation. The snake-venom protein convulxin (Cvx) binds glycoprotein (GP) VI, the platelet receptor for collagen, and triggers signal transduction. Here, the 2.7 A resolution crystal structure of Cvx is presented. In common with other members of this snake-venom protein family, Cvx is an alphabeta-heterodimer and conforms to the C-type lectin-fold topology. Comparison with other family members allows a set of Cvx residues that form a concave surface to be putatively implicated in GPVI binding. Unlike other family members, with the exception of flavocetin-A (FL-A), Cvx forms an (alphabeta)(4) tetramer. This oligomeric structure is consistent with Cvx clustering GPVI molecules on the surface of platelets and as a result promoting signal transduction activity. The Cvx structure and the location of the putative binding sites suggest a model for this multimeric signalling assembly.

Methodology employed for the structure determination of tumour necrosis factor, a case of high non-crystallographic symmetry

The structure of the protein tumour necrosis factor (TNF) was determined from crystals of space group P3(1)21 which contain six copies of the TNF monomer per crystallographic asymmetric unit [Jones, Stuart & Walker (1989). Nature (London), 338, 225-228]. The nature of these crystals (relatively high crystallographic symmetry coupled with multiple copies of the protein in the asymmetric unit) led to some peculiarly challenging problems at several points in the structure determination. In particular, (1) self-rotation function calculations failed to yield clearly interpretable solutions, (2) the analysis of difference Patterson maps for heavy-atom derivatives required the development of a Patterson search program suite GROPAT. The redundancy in the asymmetric unit allowed refinement of poor-quality isomorphous phases at 4 A resolution and phase extension from 4 to 2.9 A resolution using real-space symmetry averaging and solvent flattening in the absence of any isomorphous phase information. Despite further difficulties caused by structural differences between the six independent copies of the monomer the resultant electron density map was of high quality and proved to be easily interpretable.

Crystallization and functional analysis of a soluble deglycosylated form of the human costimulatory molecule B7‐1

The interactions of B7-1 with CD28 and CTLA-4 modulate the course of human immune responses, making B7-1 an important target for developing structure-based therapeutics. B7-1 is, however, one of the most heavily glycosylated proteins found at the leukocyte cell surface, complicating the structural analysis of this molecule. Methods for the production, crystallization and selenomethionine labelling of a soluble deglycosylated form of this molecule are described. The protein readily forms both tetragonal plate and bipyramidal crystals belonging to space groups I4(1)22, with unit-cell parameters a = b = 56.9, c = 298.7 A, and P4(1)22 (or P4(3)22), with unit-cell parameters a = b = 89.0, c = 261.9 A, respectively. The I4(1)22 and primitive crystal forms diffract to 2.7 and 3.5 A, respectively. Surface plasmon resonance-based assays indicate that the ligand-binding properties of sB7-1 are unaffected by deglycosylation. Since none of the methods relied on any special structural properties of sB7-1, it is proposed that this novel combination of procedures could in principle be adapted to the systematic analysis of many other glycoproteins of structural or functional interest.

Structure of HLA-A*0301 in complex with a peptide of proteolipid protein: insights into the role of HLA-A alleles in susceptibility to multiple sclerosis

The structure of the human major histocompatability (MHC) class I molecule HLA-A*0301 (HLA-A3) in complex with a nonameric peptide (KLIETYFSK) has been determined by X-ray crystallography to 2.7 Å resolution.

The growth and characterization of crystals of human immunodeficiency virus (HIV) reverse transcriptase

Abstract Extensive studies on the crystallization of HIV-1 reverse transcriptase (RT) have yielded several crystal forms, two of which show diffraction to minimum Bragg spacings of 6 A or better. Type 1 crystals belong to the space group P2 1 2 1 2 1 with unit cell dimensions a = 147 A , b = 190 A and c = 182 A . Crystal density measurement indicate a very high crystal solvent content of 77% consistent with the presence of two RT heterodimers (66k/51k) per crystallographic asymmetric unit. These crystals are suitable for a low resolution determination of the apoenzyme structure. The second well ordered crystal form (space group P4 2 22 with unit cell dimensions a = b = 120 A , c = 320 A ) results from the co-crystallization of RT heterodimer and a double-stranded DNA oligonucleotide. Crystal density measurements again yield a relatively high value for the solvent content (7%; one RT heterodimer per crystallographic asymmetric unit) and elemental analysis indicates that one DNA oligonucleotide is associated with each RT heterodimer. This is consistent with each heterodimer possessing a single, competent, active site.