The Vaccinia Virus DNA Helicase Structure from Combined Single-Particle Cryo-Electron Microscopy and AlphaFold2 Prediction.

Poxviruses are large DNA viruses with a linear double-stranded DNA genome circularized at the extremities. The helicase-primase D5, composed of six identical 90 kDa subunits, is required for DNA replication. D5 consists of a primase fragment flexibly attached to the hexameric C-terminal polypeptide (res. 323-785) with confirmed nucleotide hydrolase and DNA-binding activity but an elusive helicase activity. We determined its structure by single-particle cryo-electron microscopy. It displays an AAA+ helicase core flanked by N- and C-terminal domains. Model building was greatly helped by the predicted structure of D5 using AlphaFold2. The 3.9 Å structure of the N-terminal domain forms a well-defined tight ring while the resolution decreases towards the C-terminus, still allowing the fit of the predicted structure. The N-terminal domain is partially present in papillomavirus E1 and polyomavirus LTA helicases, as well as in a bacteriophage NrS-1 helicase domain, which is also closely related to the AAA+ helicase domain of D5. Using the Pfam domain database, a D5_N domain followed by DUF5906 and Pox_D5 domains could be assigned to the cryo-EM structure, providing the first 3D structures for D5_N and Pox_D5 domains. The same domain organization has been identified in a family of putative helicases from large DNA viruses, bacteriophages, and selfish DNA elements.

Solution Structure of the C-terminal Domain of A20, the Missing Brick for the Characterization of the Interface between Vaccinia Virus DNA Polymerase and its Processivity Factor.

Poxviruses are enveloped viruses with a linear, double-stranded DNA genome. Viral DNA synthesis is achieved by a functional DNA polymerase holoenzyme composed of three essential proteins. For vaccinia virus (VACV) these are E9, the catalytic subunit, a family B DNA polymerase, and the heterodimeric processivity factor formed by D4 and A20. The A20 protein links D4 to the catalytic subunit. High-resolution structures have been obtained for the VACV D4 protein in complex with an N-terminal fragment of A20 as well as for E9. In addition, biochemical studies provided evidence that a poxvirus-specific insertion (insert 3) in E9 interacts with the C-terminal residues of A20. Here, we provide solution structures of two different VACV A20 C-terminal constructs containing residues 304-426, fused at their C-terminus to either a BAP (Biotin Acceptor Peptide)-tag or a short peptide containing the helix of E9 insert 3. Together with results from titration studies, these structures shed light on the molecular interface between the catalytic subunit and the processivity factor component A20. The interface comprises hydrophobic residues conserved within the Chordopoxvirinae subfamily. Finally, we constructed a HADDOCK model of the VACV A20304-426-E9 complex, which is in excellent accordance with previous experimental data.

Intermediate-resolution crystal structure of the human adenovirus B serotype 3 fibre knob in complex with the EC2-EC3 fragment of desmoglein 2.

The cryo-electron microscopy (cryo-EM) structure of the complex between the trimeric human adenovirus B serotype 3 fibre knob and human desmoglein 2 fragments containing cadherin domains EC2 and EC3 has been published, showing 3:1 and 3:2 complexes. Here, the crystal structure determined at 4.5 Šresolution is presented with one EC2-EC3 desmoglein fragment bound per fibre knob monomer in the asymmetric unit, leading to an apparent 3:3 stoichiometry. However, in concentrated solution the 3:2 complex is predominant, as shown by small-angle X-ray scattering (SAXS), while cryo-EM at lower concentrations showed a majority of the 3:1 complex. Substitution of the calcium ions bound to the desmoglein domains by terbium ions allowed confirmation of the X-ray model using their anomalous scattering and shows that at least one binding site per cluster of calcium ions is intact and exchangeable and, combined with SAXS data, that the cadherin domains are folded even in the distal part that is invisible in the cryo-EM reconstruction.

CryoEM structure of adenovirus type 3 fibre with desmoglein 2 shows an unusual mode of receptor engagement.

Attachment of human adenovirus (HAd) to the host cell is a critical step of infection. Initial attachment occurs via the adenoviral fibre knob protein and a cellular receptor. Here we report the cryo-electron microscopy (cryo-EM) structure of a <100 kDa non-symmetrical complex comprising the trimeric HAd type 3 fibre knob (HAd3K) and human desmoglein 2 (DSG2). The structure reveals a unique stoichiometry of 1:1 and 2:1 (DSG2: knob trimer) not previously observed for other HAd-receptor complexes. We demonstrate that mutating Asp261 in the fibre knob is sufficient to totally abolish receptor binding. These data shed new light on adenovirus infection strategies and provide insights for adenoviral vector development and structure-based design.

Domain Organization of Vaccinia Virus Helicase-Primase D5.

UNLABELLED: Poxviridae are viruses with a large linear double-stranded DNA genome coding for up to 250 open reading frames and a fully cytoplasmic replication. The double-stranded DNA genome is covalently circularized at both ends. Similar structures of covalently linked extremities of the linear DNA genome are found in the African swine fever virus (asfarvirus) and in the Phycodnaviridae We are studying the machinery which replicates this peculiar genome structure. From our work with vaccinia virus, we give first insights into the overall structure and function of the essential poxvirus virus helicase-primase D5 and show that the active helicase domain of D5 builds a hexameric ring structure. This hexamer has ATPase and, more generally, nucleoside triphosphatase activities that are indistinguishable from the activities of full-length D5 and that are independent of the nature of the base. In addition, hexameric helicase domains bind tightly to single- and double-stranded DNA. Still, the monomeric D5 helicase construct truncated within the D5N domain leads to a well-defined structure, but it does not have ATPase or DNA-binding activity. This shows that the full D5N domain has to be present for hexamerization. This allowed us to assign a function to the D5N domain which is present not only in D5 but also in other viruses of the nucleocytoplasmic large DNA virus (NCLDV) clade. The primase domain and the helicase domain were structurally analyzed via a combination of small-angle X-ray scattering and, when appropriate, electron microscopy, leading to consistent low-resolution models of the different proteins. IMPORTANCE: Since the beginning of the 1980s, research on the vaccinia virus replication mechanism has basically stalled due to the absence of structural information. As a result, this important class of pathogens is less well understood than most other viruses. This lack of information concerns in general viruses of the NCLDV clade, which use a superfamily 3 helicase for replication, as do poxviruses. Here we provide for the first time information about the domain structure and DNA-binding activity of D5, the poxvirus helicase-primase. This result not only refines the current model of the poxvirus replication fork but also will lead in the long run to a structural basis for antiviral drug design.

Production and characterisation of Epstein-Barr virus helicase-primase complex and its accessory protein BBLF2/3.

The helicase-primase complex is part of the lytic DNA replication machinery of herpesviruses, but up to now, almost nothing is known about its structure. For Epstein-Barr virus it consists in the helicase BBLF4, the primase BSLF1 and the accessory protein BBLF2/3. The accessory protein shows only weak sequence homology within the herpesvirus family but may be related to an inactive B-family polymerase. BSLF1 belongs to the archaeo-eukaryotic primase family, whereas the helicase BBLF4 has been related either to Dda helicases of caudovirales or to Pif1 helicases. We produced the helicase-primase complex in insect cells using a baculovirus coding for all three proteins simultaneously. The soluble monomeric helicase-primase complex containing the three proteins with 1:1:1 stoichiometry showed ATPase activity, which is strongly stimulated in the presence of ssDNA oligomers. Furthermore, we expressed BBLF2/3 as soluble monomeric protein and performed small-angle X-ray scattering experiments which yielded an envelope whose shape is compatible with B-family polymerases.

Interaction of Ubinuclein-1, a nuclear and adhesion junction protein, with the 14-3-3 epsilon protein in epithelial cells: implication of the PKA pathway.

Ubinuclein-1 is a NACos (Nuclear and Adhesion junction Complex components) protein which shuttles between the nucleus and tight junctions, but its function in the latter is not understood. Here, by co-immunoprecipitation and confocal analysis, we show that Ubinuclein-1 interacts with the 14-3-3ɛ protein both in HT29 colon cells, and AGS gastric cells. This interaction is mediated by an Ubinuclein-1 phosphoserine motif. We show that the arginine residues (R56, R60 and R132) which form the 14-3-3ɛ ligand binding site are responsible for the binding of 14-3-3ɛ to phosphorylated Ubinuclein-1. Furthermore, we demonstrate that in vitro Ubinuclein-1 can be directly phosphorylated by cAMP-dependent protein kinase A. This in vitro phosphorylation allows binding of wildtype 14-3-3ɛ. Moreover, treatment of the cells with inhibitors of the cAMP-dependent protein kinase, KT5720 or H89, modifies the subcellular localization of Ubinuclein-1. Indeed, KT5720 and H89 greatly increase the staining of Ubinuclein-1 at the tight junctions in AGS gastric cells. In the presence of the kinase inhibitor KT5720, the amount of Ubinuclein-1 in the NP40 insoluble fraction is increased, together with actin. Moreover, treatment of the cells with KT5720 or H89 induces the concentration of Ubinuclein-1 at tricellular intersections of MDCK cells. Taken together, our findings demonstrate novel cell signaling trafficking by Ubinuclein-1 via association with 14-3-3ɛ following Ubinuclein-1 phosphorylation by the cAMP-dependent protein kinase-A.

The structural basis for the integrity of adenovirus Ad3 dodecahedron.

During the viral life cycle adenoviruses produce excess capsid proteins. Human adenovirus serotype 3 (Ad3) synthesizes predominantly an excess of free pentons, the complexes of pentameric penton base and trimeric fiber proteins, which are responsible for virus penetration. In infected cells Ad3 pentons spontaneously assemble into dodecahedral virus-like nano-particles containing twelve pentons. They also form in insect cells during expression in the baculovirus system. Similarly, in the absence of fiber protein dodecahedric particles built of 12 penton base pentamers can be produced. Both kinds of dodecahedra show remarkable efficiency of intracellular penetration and can be engineered to deliver several millions of foreign cargo molecules to a single target cell. For this reason, they are of great interest as a delivery vector. In order to successfully manipulate this potential vector for drug and/or gene delivery, an understanding of the molecular basis of vector assembly and integrity is critical. Crystallographic data in conjunction with site-directed mutagenesis and biochemical analysis provide a model for the molecular determinants of dodecamer particle assembly and the requirements for stability. The 3.8 Å crystal structure of Ad3 penton base dodecamer (Dd) shows that the dodecahedric structure is stabilized by strand-swapping between neighboring penton base molecules. Such N-terminal strand-swapping does not occur for Dd of Ad2, a serotype which does not form Dd under physiological conditions. This unique stabilization of the Ad3 dodecamer is controlled by residues 59-61 located at the site of strand switching, the residues involved in putative salt bridges between pentamers and by the disordered N-terminus (residues 1-47), as confirmed by site directed mutagenesis and biochemical analysis of mutant and wild type protein. We also provide evidence that the distal N-terminal residues are externally exposed and available for attaching cargo.

Allosteric competitive inactivation of hematopoietic CSF-1 signaling by the viral decoy receptor BARF1.

Hematopoietic human colony-stimulating factor 1 (hCSF-1) is essential for innate and adaptive immunity against viral and microbial infections and cancer. The human pathogen Epstein-Barr virus secretes the lytic-cycle protein BARF1 that neutralizes hCSF-1 to achieve immunomodulation. Here we show that BARF1 binds the dimer interface of hCSF-1 with picomolar affinity, away from the cognate receptor-binding site, to establish a long-lived complex featuring three hCSF-1 at the periphery of the BARF1 toroid. BARF1 locks dimeric hCSF-1 into an inactive conformation, rendering it unable to signal via its cognate receptor on human monocytes. This reveals a new functional role for hCSF-1 cooperativity in signaling. We propose a new viral strategy paradigm featuring an allosteric decoy receptor of the competitive type, which couples efficient sequestration and inactivation of the host growth factor to abrogate cooperative assembly of the cognate signaling complex.

The "Bridge" in the Epstein-Barr virus alkaline exonuclease protein BGLF5 contributes to shutoff activity during productive infection.

Replication of the human herpesvirus Epstein-Barr virus drastically impairs cellular protein synthesis. This shutoff phenotype results from mRNA degradation upon expression of the early lytic-phase protein BGLF5. Interestingly, BGLF5 is the viral DNase, or alkaline exonuclease, homologues of which are present throughout the herpesvirus family. During productive infection, this DNase is essential for processing and packaging of the viral genome. In contrast to this widely conserved DNase activity, shutoff is only mediated by the alkaline exonucleases of the subfamily of gammaherpesviruses. Here, we show that BGLF5 can degrade mRNAs of both cellular and viral origin, irrespective of polyadenylation. Furthermore, shutoff by BGLF5 induces nuclear relocalization of the cytosolic poly(A) binding protein. Guided by the recently resolved BGLF5 structure, mutants were generated and analyzed for functional consequences on DNase and shutoff activities. On the one hand, a point mutation destroying DNase activity also blocks RNase function, implying that both activities share a catalytic site. On the other hand, other mutations are more selective, having a more pronounced effect on either DNA degradation or shutoff. The latter results are indicative of an oligonucleotide-binding site that is partially shared by DNA and RNA. For this, the flexible "bridge" that crosses the active-site canyon of BGLF5 appears to contribute to the interaction with RNA substrates. These findings extend our understanding of the molecular basis for the shutoff function of BGLF5 that is conserved in gammaherpesviruses but not in alpha- and betaherpesviruses.

Identification of new interacting partners of the shuttling protein ubinuclein (Ubn-1).

We have previously characterized ubinuclein (Ubn-1) as a NACos (Nuclear and Adherent junction Complex components) protein which interacts with viral or cellular transcription factors and the tight junction (TJ) protein ZO-1. The purpose of the present study was to get more insights on the binding partners of Ubn-1, notably those present in the epithelial junctions. Using an in vivo assay of fluorescent protein-complementation assay (PCA), we demonstrated that the N-terminal domains of the Ubn-1 and ZO-1 proteins triggered a functional interaction inside the cell. Indeed, expression of both complementary fragments of venus fused to the N-terminal parts of Ubn-1 and ZO-1 was able to reconstitute a fluorescent venus protein. Furthermore, nuclear expression of the chimeric Ubn-1 triggered nuclear localization of the chimeric ZO-1. We could localize this interaction to the PDZ2 domain of ZO-1 using an in vitro pull-down assay. More precisely, a 184-amino acid region (from amino acids 39 to 223) at the N-terminal region of Ubn-1 was responsible for the interaction with the PDZ2 domain of ZO-1. Co-imunoprecipitation and confocal microscopy experiments also revealed the tight junction protein cingulin as a new interacting partner of Ubn-1. A proteomic approach based on mass spectrometry analysis (MS) was then undertaken to identify further binding partners of GST-Ubn-1 fusion protein in different subcellular fractions of human epithelial HT29 cells. LYRIC (Lysine-rich CEACAM1-associated protein) and RACK-1 (receptor for activated C-kinase) proteins were validated as bona fide interacting partners of Ubn-1. Altogether, these results suggest that Ubn-1 is a scaffold protein influencing protein subcellular localization and is involved in several processes such as cell-cell contact signalling or modulation of gene activity.

EBV lytic-phase protein BGLF5 contributes to TLR9 downregulation during productive infection.

Viruses use a wide range of strategies to modulate the host immune response. The human gammaherpesvirus EBV, causative agent of infectious mononucleosis and several malignant tumors, encodes proteins that subvert immune responses, notably those mediated by T cells. Less is known about EBV interference with innate immunity, more specifically at the level of TLR-mediated pathogen recognition. The viral dsDNA sensor TLR9 is expressed on B cells, a natural target of EBV infection. Here, we show that EBV particles trigger innate immune signaling pathways through TLR9. Furthermore, using an in vitro system for productive EBV infection, it has now been possible to compare the expression of TLRs by EBV(-) and EBV(+) human B cells during the latent and lytic phases of infection. Several TLRs were found to be differentially expressed either in latently EBV-infected cells or after induction of the lytic cycle. In particular, TLR9 expression was profoundly decreased at both the RNA and protein levels during productive EBV infection. We identified the EBV lytic-phase protein BGLF5 as a protein that contributes to downregulating TLR9 levels through RNA degradation. Reducing the levels of a pattern-recognition receptor capable of sensing the presence of EBV provides a mechanism by which the virus could obstruct host innate antiviral responses.

Structure of RavA MoxR AAA+ protein reveals the design principles of a molecular cage modulating the inducible lysine decarboxylase activity.

The MoxR family of AAA+ ATPases is widespread throughout bacteria and archaea but remains poorly characterized. We recently found that the Escherichia coli MoxR protein, RavA (Regulatory ATPase variant A), tightly interacts with the inducible lysine decarboxylase, LdcI/CadA, to form a unique cage-like structure. Here, we present the X-ray structure of RavA and show that the αßα and all-α subdomains in the RavA AAA+ module are arranged as in magnesium chelatases rather than as in classical AAA+ proteins. RavA structure also contains a discontinuous triple-helical domain as well as a ß-barrel-like domain forming a unique fold, which we termed the LARA domain. The LARA domain was found to mediate the interaction between RavA and LdcI. The RavA structure provides insights into how five RavA hexamers interact with two LdcI decamers to form the RavA-LdcI cage-like structure.

Inhibition of Epstein-Barr virus replication by small interfering RNA targeting the Epstein-Barr virus protease gene.

BACKGROUND: The Epstein-Barr virus (EBV) protease (PR), coded by the BVRF2 gene, is essential for the maturation of the viral capsid and viral DNA packaging during the late stage of the EBV lytic cycle. Like the other herpesvirus serine PRs, EBV PR could be a target for the inhibition of EBV replication. To date, no data have been reported on the inhibition of EBV PR messenger RNA (mRNA) by small interfering RNA (siRNA). METHODS: In this study, siRNAs targeting EBV PR were delivered to the epithelial 293 cell line stably transfected with the complete B95-8 EBV episome. EBV DNA and PR mRNA were quantified by real-time PCR in cells and supernatant, protein expression was assessed by immunoblotting, and production of EBV infectious particles in the culture medium was measured by Raji cell superinfection. RESULTS: The EBV PR mRNA within the cells was reduced by 73%, the PR protein by 35% and the amount of virus in the cell supernatant was drastically decreased by 86% or 95%, depending on the method. CONCLUSIONS: The strong effect of the siRNA targeting EBV PR on EBV replication attests to the crucial role played by EBV PR in the production of infectious particles and suggests that targeting this enzyme can be a new strategy against EBV-associated diseases where virus replication occurs.

The flexible motif V of Epstein-Barr virus deoxyuridine 5'-triphosphate pyrophosphatase is essential for catalysis.

Deoxyuridine 5'-triphosphate pyrophosphatases (dUTPases) are ubiquitous enzymes essential for hydrolysis of dUTP, thus preventing its incorporation into DNA. Although Epstein-Barr virus (EBV) dUTPase is monomeric, it has a high degree of similarity with the more frequent trimeric form of the enzyme. In both cases, the active site is composed of five conserved sequence motifs. Structural and functional studies of mutants based on the structure of EBV dUTPase gave new insight into the mechanism of the enzyme. A first mutant allowed us to exclude a role in enzymatic activity for the disulfide bridge involving the beginning of the disordered C terminus. Sequence alignments revealed two groups of dUTPases, based on the position in sequence of a conserved aspartic acid residue close to the active site. Single mutants of this residue in EBV dUTPase showed a highly impaired catalytic activity, which could be partially restored by a second mutation, making EBV dUTPase more similar to the second group of enzymes. Deletion of the flexible C-terminal tail carrying motif V resulted in a protein completely devoid of enzymatic activity, crystallizing with unhydrolyzed Mg(2+)-dUTP complex in the active site. Point mutations inside motif V highlighted the essential role of lid residue Phe(273). Magnesium appears to play a role mainly in substrate binding, since in absence of Mg(2+), the K(m) of the enzyme is reduced, whereas the k(cat) is less affected.

A bridge crosses the active-site canyon of the Epstein-Barr virus nuclease with DNase and RNase activities.

Epstein-Barr virus, a double-stranded DNA (dsDNA) virus, is a major human pathogen from the herpesvirus family. The nuclease is one of the lytic cycle proteins required for successful viral replication. In addition to the previously described endonuclease and exonuclease activities on single-stranded DNA and dsDNA substrates, we observed an RNase activity for Epstein-Barr virus nuclease in the presence of Mn(2+), giving a possible explanation for its role in host mRNA degradation. Its crystal structure shows a catalytic core of the D-(D/E)XK nuclease superfamily closely related to the exonuclease from bacteriophage lambda with a bridge across the active-site canyon. This bridge may reduce endonuclease activity, ensure processivity or play a role in strand separation of dsDNA substrates. As the DNA strand that is subject to cleavage is likely to make a sharp turn in front of the bridge, endonuclease activity on single-stranded DNA stretches appears to be possible, explaining the cleavage of circular substrates.

New insights on the role of the gamma-herpesvirus uracil-DNA glycosylase leucine loop revealed by the structure of the Epstein-Barr virus enzyme in complex with an inhibitor protein.

Epstein-Barr virus (EBV) is a human gamma-herpesvirus. Within its 86 open reading frame containing genome, two enzymes avoiding uracil incorporation into DNA can be found: uracil triphosphate hydrolase and uracil-DNA glycosylase (UNG). The latter one excises uracil bases that are due to cytosine deamination or uracil misincorporation from double-stranded DNA substrates. The EBV enzyme belongs to family 1 UNGs. We solved the three-dimensional structure of EBV UNG in complex with the uracil-DNA glycosylase inhibitor protein (Ugi) from bacteriophage PBS-2 at a resolution of 2.3 A by X-ray crystallography. The structure of EBV UNG encoded by the BKRF3 reading frame shows the excellent global structural conservation within the solved examples of family 1 enzymes. Four out of the five catalytic motifs are completely conserved, whereas the fifth one, the leucine loop, carries a seven residue insertion. Despite this insertion, catalytic constants of EBV UNG are similar to those of other UNGs. Modelling of the EBV UNG-DNA complex shows that the longer leucine loop still contacts DNA and is likely to fulfil its role of DNA binding and deformation differently than the enzymes with previously solved structures. We could show that despite the evolutionary distance of EBV UNG from the natural host protein, bacteriophage Ugi binds with an inhibitory constant of 8 nM to UNG. This is due to an excellent specificity of Ugi for conserved elements of UNG, four of them corresponding to catalytic motifs and a fifth one corresponding to an important beta-turn structuring the catalytic site.

Kinetics, inhibition and oligomerization of Epstein-Barr virus protease.

Epstein-Barr virus (EBV) is an omnipresent human virus causing infectious mononucleosis and EBV associated cancers. Its protease is a possible target for antiviral therapy. We studied its dimerization and enzyme kinetics with two enzyme assays based either on the release of paranitroaniline or 7-amino-4-methylcoumarin from labeled pentapeptide (Ac-KLVQA) substrates. The protease is in a monomer-dimer equilibrium where only dimers are active. In absence of citrate the K(d) is 20 microM and drops to 0.2 microM in presence of 0.5M citrate. Citrate increases additionally the activity of the catalytic sites. The inhibitory constants of different substrate derived peptides and alpha-keto-amide based inhibitors, which have at best a K(i) of 4 microM, have also been evaluated.

Structural genomics of the Epstein-Barr virus.

Epstein-Barr virus is a herpesvirus that causes infectious mononucleosis, carcinomas and immunoproliferative disease. Its genome encodes 86 proteins, which provided targets for a structural genomics project. After updating the annotation of the genome, 23 open reading frames were chosen for expression in Escherichia coli, initially selecting for those with known enzyme activity and then supplementing this set based on a series of predicted properties, in particular secondary structure. The major obstacle turned out to be poor expression and low solubility. Surprisingly, this could not be overcome by modifications of the constructs, changes of expression temperature or strain or renaturation. Of the eight soluble proteins, five were crystallized using robotic nanolitre-drop crystallization trials, which led to four solved structures. Although these results depended on individual treatment rather than standardized protocols, a high-throughput miniaturized crystallization screening protocol was a key component of success with these difficult proteins.

Structure of the Epstein-Barr virus oncogene BARF1.

The Epstein-Barr virus is a human gamma-herpesvirus that persistently infects more than 90% of the human population. It is associated with numerous epithelial cancers, principally undifferentiated nasopharyngeal carcinoma and gastric carcinoma. The BARF1 gene is expressed in a high proportion of these cancers. An oncogenic, mitogenic and immortalizing activity of the BARF1 protein has been shown. We solved the structure of the secreted BARF1 glycoprotein expressed in a human cell line by X-ray crystallography at a resolution of 2.3A. The BARF1 protein consists of two immunoglobulin (Ig)-like domains. The N-terminal domain belongs to the subfamily of variable domains whereas the C-terminal one is related to a constant Ig-domain. BARF1 shows an unusual hexamerisation involving two principal contacts, one between the C-terminal domains and one between the N-terminal domains. The C-terminal contact with an uncommonly large contact surface extends the beta-sandwich of the Ig-domain through the second molecule. The N-terminal contact involves Ig-domains with an unusual relative orientation but with a more classical contact surface with a size in the range of dimer interactions of Ig-domains. The structure of BARF1 is most closely related to CD80 or B7-1, a co-stimulatory molecule present on antigen presenting cells, from which BARF1 must have been derived during evolution. Still, domain orientation and oligomerization differ between BARF1 and CD80. It had been shown that BARF1 binds to hCSF-1, the human colony-stimulating factor 1, but this interaction has to be principally different from the one between CSF-1 and CSF-1 receptor.