The crystal structure of the costimulatory OX40-OX40L complex.

OX40 is a T cell costimulator activated by OX40L. Blockade of the OX40L-OX40 interaction has ameliorative effects in animal models of T cell pathologies. In order to better understand the interaction between OX40 and OX40L, we have determined the crystal structure of murine OX40L and of the human OX40-OX40L complex at 1.45 and 2.4 A, respectively. These structures show that OX40L is an unusually small member of the tumor necrosis factor superfamily (TNFSF). The arrangement of the OX40L protomers forming the functional trimer is atypical and differs from that of other members by a 15 degrees rotation of each protomer with respect to the trimer axis, resulting in an open assembly. Site-directed changes of the interfacial residues of OX40L suggest this interface lacks a single "hot spot" and that instead, binding energy is dispersed over at least two distinct areas. These structures demonstrate the structural plasticity of TNFSF members and their interactions with receptors.

Activation of the proapoptotic death receptor DR5 by oligomeric peptide and antibody agonists.

The cell-extrinsic apoptotic pathway triggers programmed cell death in response to certain ligands that bind to cell-surface death receptors. Apoptosis is essential for normal development and homeostasis in metazoans, and furthermore, selective activation of the cell-extrinsic pathway in tumor cells holds considerable promise for cancer therapy. We used phage display to identify peptides and synthetic antibodies that specifically bind to the human proapoptotic death receptor DR5. Despite great differences in overall size and structure, the DR5-binding peptides and antibodies shared a tripeptide motif, which was conserved within a disulfide-constrained loop of the peptides and the third complementarity determining region of the antibody heavy chains. The X-ray crystal structure of an antibody in complex with DR5 revealed that the tripeptide motif is buried at the core of the interface, confirming its central role in antigen recognition. We found that certain peptides and antibodies exhibited potent proapoptotic activity against DR5-expressing SK-MES-1 lung carcinoma cells. These phage-derived ligands may be useful for elucidating DR5 activation at the molecular level and for creating synthetic agonists of proapoptotic death receptors.

Molecular recognition by a binary code.

Functional antibodies were obtained from a library of antigen-binding sites generated by a binary code restricted to tyrosine and serine. An antibody raised against human vascular endothelial growth factor recognized the antigen with high affinity (K(D)=60 nM) and high specificity in cell-based assays. The crystal structure of another antigen binding fragment in complex with its antigen (human death receptor DR5) revealed the structural basis for this minimalist mode of molecular recognition. Natural antigen-binding sites are enriched for tyrosine and serine, and we show that these amino acid residues are intrinsically well suited for molecular recognition. Furthermore, these results demonstrate that molecular recognition can evolve from even the simplest chemical diversity.

The crystal structures of EDA-A1 and EDA-A2: splice variants with distinct receptor specificity.

EDA is a tumor necrosis factor family member involved in ectodermal development. Splice variants EDA-A1 and EDA-A2 differ only by the presence of Glu 308 and Val 309 in the expected receptor binding region of EDA-A1 but not EDA-A2. This two amino acid difference functions as a switch controlling receptor specificity. EDA-A1 binds only to EDAR, while EDA-A2 is specific for XEDAR. In order to understand the structural basis of this switch, we determined the X-ray crystal structures of the TNF domain of both EDA-A1 and EDA-A2 at 2.3 A and 2.2 A, respectively. While the backbone conformation around the splice difference is similar in both isoforms, the conformation of the following loop, the surface charge, and the shape of the expected receptor binding site differ significantly.

The crystal structure of a proliferation-inducing ligand, APRIL.

A proliferation-inducing ligand (APRIL) is a TNF-like cytokine that stimulates tumor cell growth. Within the TNF ligand superfamily, APRIL is most similar to B-cell activation factor (BAFF) with which it shares 30% sequence identity. APRIL binds the receptors B-cell maturation antigen (BCMA) and TACI with high affinity; both of these receptors have also been shown to bind BAFF, although BCMA has significantly higher affinity for APRIL than BAFF. Determination of the crystal structure of APRIL from three crystallization conditions at resolutions of 1.8-2.4A over a pH range from 5.0 to 8.5 reveals a compact trimeric ligand with a backbone fold very similar to that of BAFF (1.1A RMSD over 122 structurally equivalent Calpha atoms), with the exception of differences in the AA' and DE loop regions. Whereas BAFF has been shown to form 20-trimer assemblies under certain conditions, the molecular determinants required for BAFF-like assemblies are absent in the APRIL structure. No crystal packing suggestive of the formation of higher-order assemblies is seen in any of the crystal forms nor does the structure vary significantly between pH 5.0 and 8.5. Modeling of the APRIL-BCMA complex shows the resulting interface is in agreement with mutagenesis data.

Preparation of distinct ubiquitin chain reagents of high purity and yield.

The complexity of protein ubiquitination signals derives largely from the variety of polyubiquitin linkage types that can modify a target protein, each imparting distinct functional consequences. Free ubiquitin chains of uniform linkages and length are important tools in understanding how ubiquitin-binding proteins specifically recognize these different polyubiquitin modifications. While some free ubiquitin chain species are commercially available, mutational analyses and labeling schemes are limited to select, marketed stocks. Furthermore, the multimilligram quantities of material required for detailed biophysical and/or structural studies often makes these reagents cost prohibitive. To address these limitations, we have optimized known methods for the synthesis and purification of linear, K11-, K48-, and K63-linked ubiquitin dimers, trimers, and tetramers on a preparative scale. The high purity and relatively high yield of these proteins readily enables material-intensive experiments and provides flexibility for engineering specialized ubiquitin chain reagents, such as fluorescently labeled chains of discrete lengths.

Synthetic anti-BR3 antibodies that mimic BAFF binding and target both human and murine B cells.

BR3, which is expressed on all mature B cells, is a specific receptor for the B-cell survival and maturation factor BAFF (B-cell-activating factor belonging to the tumor necrosis factor [TNF] family). In order to investigate the consequences of targeting BR3 in murine models and to assess the potential of BR3 antibodies as human therapeutics, synthetic antibody phage libraries were employed to identify BAFF-blocking antibodies cross-reactive to murine and human BR3, which share 52% identity in their extracellular domains. We found an antibody, CB1, which exhibits muM affinity for murine BR3 and very weak affinity for the human receptor. CB3s, an affinity-matured variant of CB1, has sub-nM affinity for BR3 from both species. Alanine scanning and crystallographic structural analysis of the CB3s/BR3 complex reveal that CB3s mimics BAFF by interacting with a similar region of the BR3 surface. Despite this similarity in binding epitopes, CB1 variants antagonize BAFF-dependent human B-cell proliferation in vitro and are effective at reducing murine B-cell populations in vivo, showing significant promise as therapeutics for human B-cell-mediated diseases.

Attenuating lymphocyte activity: the crystal structure of the BTLA-HVEM complex.

Five CD28-like proteins exert positive or negative effects on immune cells. Only four of these five receptors interact with members of the B7 family. The exception is BTLA (B and T lymphocyte attenuator), which instead interacts with the tumor necrosis factor receptor superfamily member HVEM (herpes virus entry mediator). To better understand this interaction, we determined the 2.8-A crystal structure of the BTLA-HVEM complex. This structure shows that BTLA binds the N-terminal cysteine-rich domain of HVEM and employs a unique binding surface compared with other CD28-like receptors. Moreover, the structure shows that BTLA recognizes the same surface on HVEM as gD (herpes virus glycoprotein D) and utilizes a similar binding motif. Light scattering analysis demonstrates that the extracellular domain of BTLA is monomeric and that BTLA and HVEM form a 1:1 complex. Alanine-scanning mutagenesis of HVEM was used to further define critical binding residues. Finally, BTLA adopts an immunoglobulin I-set fold. Despite structural similarities to other CD28-like members, BTLA represents a unique co-receptor.

Functional consequences of a gene duplication and fusion event in an arginine kinase.

Arginine kinase (AK) from the foot of the razor clam Ensis directus consists of two full-length AK domains, denoted D1 and D2, fused in a single polypeptide chain. The full-length cDNA for Ensis AK was obtained and its deduced amino acid sequence was analyzed in the context of the X-ray crystal structure of a typical, monomeric AK. Both domains of Ensis AK contain most of the residues currently thought to be critical in catalysis, suggesting that both AK domains are catalytically competent. The full-length Ensis AK, a D2-NusA-His-tag fusion protein and a D2-truncated AK (enterokinase cleavage product of the fusion protein) were expressed in Escherichia coli and purified. All recombinant AK constructs displayed high enzyme activity. Attempts at expressing active D1 alone, D2 alone or a D1-NusA-His-tag fusion protein were unsuccessful. The catalytic properties of the active proteins were compared with the corresponding properties of recombinant AK from the horseshoe crab Limulus polyphemus, which is a typical monomeric AK. In contrast to expectations, the kinetic results strongly suggest that Ensis AK has only one active domain, namely D2. The K(cat) values for all Ensis constructs were roughly twice that of typical AKs, indicating higher overall catalytic throughput at the competent active site. Furthermore, both the full-length and truncated D2 Ensis AKs showed no synergism of substrate binding unlike typical AKs. The D2-NusA-His-tag fusion construct actually displayed negative synergism of substrate binding, which means that, in effect, the first substrate bound acts as a competitive inhibitor of the second. The conservation of the structure of the apparently inactive D1 may be related to constraints imposed by structural changes that could potentially impact substrate binding in D2 and/or possibly influence the proper folding of the enzyme during synthesis. Overall, the results from the present study indicate that the AK contiguous dimer from Ensis directus functions with activity in only the second domain. Although lacking activity in D1, D2 appears to compensate by having a higher intrinsic catalytic throughput than typical 40-kDa monomeric AKs.

Evolution and divergence of the genes for cytoplasmic, mitochondrial, and flagellar creatine kinases.

Creatine kinase (CK) plays a central role in energy homeostasis in cells that display high and variable rates of energy turnover. A number of CK genes exist, each being targeted to particular intracellular compartments. In the vertebrates, two genes code for proteins which form homo- and heterodimers targeted to the cytoplasm, while two additional genes code for primarily octameric proteins targeted to the mitochondrial intermembrane space. Yet another gene is present in certain groups which codes for three fused, complete CK domains and is typically targeted to the flagellar membrane of primitive-type spermatozoa. CK is widely distributed in protochordates and both protostome and deuterostome invertebrate groups. The evolutionary relationships of these CK genes have not been fully elucidated. The present communication reports new cDNA-derived deduced amino acid sequences for four cytoplasmic and three mitochondrial CKs and one flagellar CK from lophotrochozoan, protostome invertebrates as well as a new cytoplasmic CK sequence from a protochordate tunicate. These new sequences, coupled with available sequences in the databases and sequences extracted from genome sequencing projects, provide revealing insights into the evolution and divergence of CK genes. Phylogenetic analyses showed that single cytoplasmic, mitochondrial, and flagellar CK genes were present prior to the divergence of the protostomes and deuterostomes. The flagellar CK gene may have evolved within the cytoplasmic gene clade, although the evidence is somewhat equivocal. The two cytoplasmic genes in the vertebrates, and most likely the two mitochondrial genes, evolved after the divergence of the craniates from the protochordates. Comparison of the structure of the genes for selected cytoplasmic, mitochondrial, and flagellar CKs revealed two identical intron boundaries, further reinforcing the notion of a common evolutionary origin, but also showed patterns of changes in structure consistent with each gene type. These studies show that the cytoplasmic, mitochondrial, and flagellar CK genes are rather ancient and that there has been a systematic pattern of duplication and divergence consistent with changing nature of energy demands and physicochemical environment in the cells where they are expressed.