Disulphide-linked Homodimer Research Articles

Functional identification and characterization of IpMSTNa, a novel orthologous myostatin (MSTN) gene in channel catfish Ictalurus punctatus

Channel catfish (Ictalurus punctatus) are one of the most important commercial freshwater fish in the world. China has been the major producer and consumer of channel catfish following the rapid development in the past three decades. In the present study, a novel orthologous myostatin gene, IpMSTNa, of channel catfish was identified based on homology cloning and genome locating. Multiple sequence alignments and gene structure analyses showed that the IpMSTNa gene and its deduced protein presented similar architectures to other known vertebrates. Phylogenetic and synteny analyses indicated that IpMSTNa belongs to MSTN1 orthologues. Pro-IpMSTNa protein is a typical disulphide-linked homodimer, with each chain containing an N-terminal pro-domain and a C-terminal unmatured GF domain, while pro-IpMSTNa present some significant differences in secondary structure and three-dimensional substances with pro-IpMSTNb. Relative expression level of the IpMSTNa gene upregulated rapidly and decreased dramatically during the embryonic and larval developmental stages, respectively. In addition, IpMSTNa displayed remarkably higher expression at most developmental stages compared to IpMSTNb. Tissue distribution analysis indicated that the IpMSTNa gene had a significantly higher level of expression than IpMSTNb in all selected tissues, with abundantly greater expression in the liver, muscle, gill and spleen, and moderately greater expression in the kidney, intestine, and head kidney. ISH analysis demonstrated that the expression signals of IpMSTNa and IpMSTNb at the selected developmental stages are consistent to qRT-PCR tests. Our study suggested that the IpMSTNa gene may have more biological functions, which have yet to be determined compared to the IpMSTNb gene.

Restoration of glycoprotein Erns dimerization via pseudoreversion partially restores virulence of classical swine fever virus.

The classical swine fever virus (CSFV) represents one of the most important pathogens of swine. The CSFV glycoprotein Erns is an essential structural protein and an important virulence factor. The latter is dependent on the RNase activity of this envelope protein and, most likely, its secretion from the infected cell. A further important feature with regard to its function as a virulence factor is the formation of disulfide-linked Erns homodimers that are found in virus-infected cells and virions. Mutant CSFV lacking cysteine (Cys) 171, the residue responsible for intermolecular disulfide bond formation, were found to be attenuated in pigs (Tews BA, Schürmann EM, Meyers G. J Virol 2009;83:4823-4834). In the course of an animal experiment with such a dimerization-negative CSFV mutant, viruses were reisolated from pigs that contained a mutation of serine (Ser) 209 to Cys. This mutation restored the ability to form disulphide-linked Erns homodimers. In transient expression studies Erns mutants carrying the S209C change were found to form homodimers with about wt efficiency. Also the secretion level of the mutated proteins was equivalent to that of wt Erns. Virus mutants containing the Cys171Ser/Ser209Cys configuration exhibited wt growth rates and increased virulence when compared with the Cys171Ser mutant. These results provide further support for the connection between CSFV virulence and Erns dimerization.

Transmembrane features governing Fc receptor CD16A assembly with CD16A signaling adaptor molecules

Many activating immunoreceptors associate with signaling adaptor molecules like FcεR1γ or CD247. FcεR1γ and CD247 share high sequence homology and form disulphide-linked homodimers that contain a pair of acidic aspartic acid residues in their transmembrane (TM) domains that mediate assembly, via interaction with an arginine residue at a similar register to these aspartic acids, with the activating immunoreceptors. However, this model cannot hold true for receptors like CD16A, whose TM domains do not contain basic residues. We have carried out an extensive site-directed mutagenesis analysis of the CD16A receptor complex and now report that the association of receptor with the signaling adaptor depends on a network of polar and aromatic residues along the length of the TM domain. Molecular modeling indicates that CD16A TM residues F202, D205, and T206 form the core of the membrane-embedded trimeric interface by establishing highly favorable contacts to the signaling modules through rearrangement of a hydrogen bond network previously identified in the CD247 TM dimer solution NMR structure. Strikingly, the amino acid D205 also regulates the turnover and surface expression of CD16A in the absence of FcεR1γ or CD247. Modeling studies indicate that similar features underlie the association of other activating immune receptors, including CD64 and FcεR1α, with signaling adaptor molecules, and we confirm experimentally that equivalent F, D, and T residues in the TM domain of FcεR1α markedly influence the biology of this receptor and its association with FcεR1γ.

From bones to blood pressure, developing novel biologic approaches targeting the osteoprotegein pathway for pulmonary vascular disease.

Osteoprotegerin (tnfsf11b, OPG) is a soluble member of the TNF superfamily originally described as an important regulator of osteoclastogenesis almost 20 years ago. OPG is a heparin-binding secreted glycoprotein that exists as a 55–62 kDa monomer or a 110–120 kDa disulphide-linked homodimer. Acting as a soluble decoy receptor for RANKL, OPG actively regulates RANK signalling, and thereby osteoclastogenesis. OPG has subsequently been shown to also be a decoy receptor TNF related apoptosis inducing-ligand (tnfsf10, TRAIL, Apo2L). TRAIL is a type II transmembrane protein that is widely expressed in a variety of human tissues, including the spleen, lung, and prostate. Through binding to TRAIL, OPG can inhibit TRAIL-induced apoptosis of cancer cells. More recently, OPG has been demonstrated to be secreted by, and influence, vascular smooth muscle cells phenotype particularly related to vascular calcification and pulmonary vascular remodelling. In pulmonary artery smooth muscle cell (PASMC) suppression of BMP, induction of 5-HT and IL-1 signalling have been shown to stimulate the release of OPG in vitro, which causes cell migration and proliferation. Patients with idiopathic PAH (IPAH) demonstrate increased circulating and tissue levels of OPG, and circulating serum levels predict survival. In pre-clinical models, OPG levels correlate with disease severity. Since OPG is a naturally circulating protein, we are investigating the potential of novel biologic antibody therapies to rescue PAH phenotype in disease models. Further pre-clinical and mechanistic data are forthcoming, but we believe current published data identify OPG as an exciting and novel therapeutic target in PAH.

Site-specific N-glycosylation analysis of human factor XI: Identification of a noncanonical NXC glycosite.

Human factor XI (hFXI) is a 160-kDa disulphide-linked homodimer zymogen involved in the coagulation cascade. Its deficiency results in bleeding diathesis referred to as hemophilia C. hFXI bears five N-glycosylation consensus sites per monomer, N72 , N108 , N335 on the heavy chain and N432 , N473 on the light chain. This study reports the first in-depth glycosylation analysis of hFXI based on advanced MS approaches. Hydrophilic interaction LC and MS characterization and quantification of the N-glycans showed that the two major forms are complex biantennary mono-α2,6-sialylated (A2 S1 , 20%) and bis-α2,6-sialylated structures (A2 S2 , 66%). Minor triantennary structures (A3 S3 F, ∼1.5%; A3 S3 , ∼2%) were also identified. MS analyses of intact hFXI revealed full occupation of two of the three heavy-chain glycosites and almost full-site occupancy of the light chain. Analysis of hFXI glycopeptides by LC-MS/MS enabled site-specific glycan profiling and occupancy. It was evidenced that N335 was not glycosylated and that N72 and N108 were fully occupied, whereas N432 and N473 were occupied at about 92 and 95%, respectively. We also identified a new glycosite of the noncanonical format NXC at N145 , occupied at around 5%. These data provide valuable structural information useful to understand the potential roles of N-glycosylation on hFXI function and could serve as a structural reference.

Glycotyping of Trypanosoma brucei variant surface glycoprotein MITat1.8

Following a switch from variant surface glycoprotein MITat1.4 to variant surface glycoprotein MITat1.8 expression by Lister strain 427 Trypanosoma brucei brucei parasites, the latter uncharacterized variant surface glycoprotein was analysed. Variant surface glycoprotein MITat1.8 was found to be a disulphide-linked homodimer, containing a complex N-linked glycan at Asn58 and a glycosylphosphatidylinositol membrane anchor attached to Asp419. Mass spectrometric analyses demonstrated that the N-glycan is exclusively Galβ1-4GlcNAcβ1-2Manα1-3(Galβ1-4GlcNAcβ1-2Manα1-6)Manβ1-4GlcNAcβ1-4GlcNAc and that the conserved Man3GlcN-myo-inositol glycosylphosphatidylinositol anchor glycan core is substituted with an average of 4 hexose, most likely galactose, residues. The presence of a complex N-glycan at Asn58 is consistent with the relatively acidic environment of the Asn58 N-glycosylation sequon, that predicts N-glycosylation by T. brucei oligosaccharyltransferase TbSTT3A with a Man5GlcNAc2 structure destined for processing to a paucimannose and/or complex N-glycan (Izquierdo L, Schulz B, Rodrigues JA et al. EMBO J 2009;28:2650–61 [12]).

A truncated soluble epidermal growth factor receptor-Fc fusion ligand trap displays anti-tumour activity in vivo

A number of therapeutic strategies including small molecule tyrosine kinase inhibitors and monoclonal antibodies have been developed to target the epidermal growth factor receptor (EGFR) signalling axis for the treatment of cancer. To date, the focus of therapeutic intervention has been the EGFR itself. In the current study, we have assembled and expressed in mammalian cells a soluble, EGFR ligand trap comprising the first 501 amino acids of the mature EGFR sequence fused in-frame with a human IgG Fc domain. The fusion protein, designated sEGFR501.Fc, was secreted as a 220 kDa disulphide-linked homodimer that exhibited high affinity (0.4–8 nM) in competition assays for a number of EGFR ligands including EGF and transforming growth factor-α (TGF-α). sEGFR501.Fc inhibited EGF-stimulated tyrosine phosphorylation of the EGFR of the lung cancer cell lines A549 and H1437, and inhibited and blocked the proliferation of H1437 cells. Administration of sEGFR501.Fc to mice bearing human tumour xenografts derived from A431 (epidermoid carcinoma) and DU145 (androgen-independent prostate cancer) tumour cell lines resulted in modest retardation of tumour growth. These results provide proof-in-principle that using high affinity soluble receptors is a viable method for inhibiting multi-ligand systems, and the impetus to optimize this approach and develop reagents with greater affinity and broader specificity.

Effect of structural parameters of peptides on dimer formation and highly oxidized side products in the oxidation of thiols of linear analogues of human β‐defensin 3 by DMSO

The purpose of this study was to examine the effects of structural parameters of peptides on their oxidation by DMSO, including location of cysteine, effect of adjunct group participation, molecular hydrophobicity, steric hindrance or the accessibility of thiol group and peptide conformation, on oxidation rates, dimer formation and associated side products. We designed and synthesized two series of linear cysteine-containing analogues of human beta-defensin 3 (the C1-peptides with cysteine at the N-terminus residue 1, the C29-peptides with cysteine located at residue 29 in the centre of peptide), which were used for preparation of disulphide-linked homodimers. HPLC-ESI-MS was used to monitor the oxidation process and to characterize the molecular weights of dimers and side products of high oxidation. The formations of dimers and side products were dependent on the position of cysteines. Hydrophobicity generally rendered the thiol groups less accessible and hence exposed them to slow oxidation to form dimers (or even fail to form dimers during the timescale of observation). Molecular dynamics simulations showed that the exposure of cysteines (and sulphurs) of the C1-peptides was much larger than for the C29-peptides. The larger hydrophobic side chains tended to enable clustering of the side chains that sequester cysteine, particularly in the C29-peptides, which provided a molecular explanation for the observed trends in oxidation rates. Together with molecular modelling, we propose a reaction mechanism to elucidate the oxidation results of these peptides.

Molecular properties of the partially SDS-resistant lectin from the albumen gland of Helix pomatia and demonstration of lectin-related molecules on the surface of H. pomatia haemocytes

Lectins (agglutinins) are components of the immunobiological recognition system of vertebrates and invertebrates. The present study focused on the molecular properties of the agglutinin from the albumen gland of Helix pomatia (HPA) and on the occurrence of lectin-related molecules on the surface of H. pomatia haemocytes. According to the current model (Hammarstrom et al., 1972, Scandinavian Journal of Immunology, 1: 259-301), the hexameric HPA of about 79 kDa is composed of three non-covalently associated dimers (26 kDa), each consisting of two disulphide-bridged 13 kDa monomers. However, on native-gradient polyacrylamide gel electrophoresis (PAGE), we obtained high molecular weight bands representing lectin polymers. The stepwise dissociation of these was achieved by incubation with SDS at temperatures from 20 to 408C (1 h) and at 1008C (10 min). The results obtained on SDS-PAGE included the occurrence of partially SDS-resistant hexamers of about 66 kDa, of two dimer bands of 22 and 19 kDa, and of two minor heteromonomer fractions. Complete dissociation into heteromonomers of 13 and 11 kDa was achieved by boiling the lectin (10 min) with SDS under reducing conditions. For native lectin molecules, both monomers occurred as disulphide-linked homodimers. Monomers or dimers electroeluted from an SDS -gel, reassociated to SDS-resistant oligomers upon re-electrophoresis. Finally, molecules antigenetically related to the lectin were extracted from the membrane of H. pomatia haemocytes. Anti-HPA antibodies recognized peptides with an apparent molecular weight of about 30 and 56 kDa, which were shown to represent cell-surface molecules.

High allelic polymorphism, moderate sequence diversity and diversifying selection for B-NK but not B-lec, the pair of lectin-like receptor genes in the chicken MHC

We previously characterised the C-type lectin-like receptor genes B-NK and B-lec, located next to each other in opposite orientations in the chicken major histocompatibility complex (MHC). We showed that B-NK is an inhibitory receptor expressed on natural killer cells, whereas B-lec is an activation-induced receptor with a broader expression pattern. It is interesting to note that the chicken MHC has been linked with resistance or susceptibility to Marek's disease virus (MDV), an oncogenic herpes virus. Recent reports show that the C-type lectin-like receptors in mouse and rat (Ly49H, NKR-P1 and Clr) are associated with resistance to another herpesvirus, cytomegalovirus (CMV). Therefore, B-NK and B-lec are potential candidate genes for the MHC-mediated resistance to MDV. In this paper, we report that both genes encode glycosylated type II membrane proteins that form disulphide-linked homodimers. The gene sequences from nine lines of domestic chicken representing seven haplotypes show that B-lec is well conserved between the different haplotypes, apparently under purifying selection. In contrast, B-NK has high allelic polymorphism and moderate sequence diversity, with 21 nucleotide changes in the complementary deoxyribonucleic acids (cDNAs) resulting in 20 amino acid substitutions. The allelic variations include substitutions, an indel and loss/gain of three predicted N-linked glycosylation sites. Strikingly, there is as much as 7% divergence between protein sequences of B-NK from different haplotypes, greater than the difference observed between the highly polymorphic human KIR NK receptors. Analysis of ds and dn reveal evidence of strong positive selection for B-NK to be polymorphic at the protein level, and modelling demonstrates significant variation between haplotypes in the predicted ligand binding face of B-NK.

A homodimeric complex of HLA‐G on normal trophoblast cells modulates antigen‐presenting cells via LILRB1

In healthy individuals, the non-classical MHC molecule HLA-G is only expressed on fetal trophoblast cells that invade the decidua during placentation. We show that a significant proportion of HLA-G at the surface of normal human trophoblast cells is present as a disulphide-linked homodimer of the conventional β2m-associated HLA-I complex. HLA-G is a ligand for leukocyte immunoglobulin-like receptors (LILR), which bind much more efficiently to dimeric HLA-G than to conventional HLA-I molecules. We find that a LILRB1-Fc fusion protein preferentially binds the dimeric form of HLA-G on trophoblast cells. We detect LILRB1 expression on decidual myelomonocytic cells; therefore, trophoblast HLA-G may modulate the function of these cells. Co-culture with HLA-G+ cells does not inhibit monocyte-derived dendritic cell up-regulation of HLA-DR and costimulatory molecules on maturation, but did increase production of IL-6 and IL-10. Furthermore, proliferation of allogeneic lymphocytes was inhibited by HLA-G binding to LILRB1/2 on responding antigen-presenting cells (APC). As HLA-G is the only HLA-I molecule that forms β2m-associated dimers with increased avidity for LILRB1, this interaction could represent a placental-specific signal to decidual APC. We suggest that the placenta is modulating maternal immune responses locally in the uterus through HLA-G, a trophoblast-specific, monomorphic signal present in almost every pregnancy.See accompanying commentary: http://dx.doi.org/10.1002/eji.200737515

Structural Model of Human Endoglin, a Transmembrane Receptor Responsible for Hereditary Hemorrhagic Telangiectasia

Endoglin is a type I membrane protein expressed as a disulphide-linked homodimer on human vascular endothelial cells whose haploinsufficiency is responsible for the dominant vascular dysplasia known as hereditary hemorrhagic telangiectasia (HHT). Structurally, endoglin belongs to the zona pellucida (ZP) family of proteins that share a ZP domain of ∼260 amino acid residues at their extracellular region. Endoglin is a component of the TGF-β receptor complex, interacts with the TGF-β signalling receptors types I and II, and modulates cellular responses to TGF-β. Here, we have determined for the first time the three-dimensional structure of the ∼140 kDa extracellular domain of endoglin at 25 Å resolution, using single-particle electron microscopy (EM). This reconstruction provides the general architecture of endoglin, which arranges as a dome made of antiparallel oriented monomers enclosing a cavity at one end. A high-resolution structure of endoglin has also been modelled de novo and found to be consistent with the experimental reconstruction. Each subunit comprises three well-defined domains, two of them corresponding to ZP regions, organised into an open U-shaped monomer. This domain arrangement was found to closely resemble the overall structure derived experimentally and the three modelled de novo domains were tentatively assigned to the domains observed in the EM reconstruction. This molecular model was further tested by tagging endoglin's C terminus with an IgG Fc fragment visible after 3D reconstruction of the labelled protein. Combined, these data provide the structural framework to interpret endoglin's functional domains and mutations found in HHT patients.

Expression of a Dromedary Heavy Chain-Only Antibody and B Cell Development in the Mouse

In mature B cells of mice and most mammals, cellular release of single H chain Abs without L chains is prevented by H chain association with Ig-specific chaperons in the endoplasmic reticulum. In precursor B cells, however, surface expression of mu-H chain in the absence of surrogate and conventional L chain has been identified. Despite this, Ag-specific single H chain Ig repertoires, using mu-, gamma-, epsilon-, or alpha-H chains found in conventional Abs, are not produced. Moreover, removal of H chain or, separately, L chain (kappa/lambda) locus core sequences by gene targeting has prevented B cell development. In contrast, H chain-only Abs are produced abundantly in Camelidae as H2 IgG without the C(H)1 domain. To test whether H chain Abs can be produced in mice, and to investigate how their expression affects B cell development, we introduced a rearranged dromedary gamma2a H chain into the mouse germline. The dromedary transgene was expressed as a naturally occurring Ag-specific disulphide-linked homodimer, which showed that B cell development can be instigated by expression of single H chains without L chains. Lymphocyte development and B cell proliferation was accomplished despite the absence of L chain from the BCR complex. Endogenous Ig could not be detected, although V(D)J recombination and IgH/L transcription was unaltered. Furthermore, crossing the dromedary H chain mice with mice devoid of all C genes demonstrated without a doubt that a H chain-only Ab can facilitate B cell development independent of endogenous Ig expression, such as mu- or delta-H chain, at early developmental stages.

Structural and functional aspects of the Ly49 natural killer cell receptors

Natural killer cells are part of the first line of innate immune defence against virus-infected cells and cancer cells in the vertebrate immune system. They are called 'natural' killers because, unlike cytotoxic T cells, they do not require a previous challenge and preactivation to become active. The Ly49 NK receptors are type II transmembrane glycoproteins, structurally characterized as disulphide-linked homodimers. They share extensive homology with C-type lectins, and they are encoded by a multigene family that in mice maps on chromosome 6. A fine balance between inhibitory and activating signals regulates the function of NK cells. Inhibitory Ly49 molecules bind primarily MHC class I ligands, whereas the ligands for activating Ly49 molecules may include MHC class I, but also interestingly MHC class I-like molecules expressed by viruses, as is the case for Ly49H, which binds the m157 gene product of murine cytomegalovirus. In this study, we review the function and X-ray crystal structure of the Ly49 NK cell receptors hitherto determined (Ly49A, Ly49C and Ly49I), and the structural features of the Ly49/MHC class I interaction as revealed by the X-ray crystal structures of Ly49A/H-2Dd and the recently determined Ly49C/H-2Kb.

Molecular cloning and biochemical characterization of rabbit factor XI

Human factor XI, a plasma glycoprotein required for normal haemostasis, is a homodimer (160 kDa) formed by a single interchain disulphide bond linking the Cys-321 of each Apple 4 domain. Bovine, porcine and murine factor XI are also disulphide-linked homodimers. Rabbit factor XI, however, is an 80 kDa polypeptide on non-reducing SDS/PAGE, suggesting that rabbit factor XI exists and functions physiologically either as a monomer, as does prekallikrein, a structural homologue to factor XI, or as a non-covalent homodimer. We have investigated the structure and function of rabbit factor XI to gain insight into the relation between homodimeric structure and factor XI function. Characterization of the cDNA sequence of rabbit factor XI and its amino acid translation revealed that in the rabbit protein a His residue replaces the Cys-321 that forms the interchain disulphide linkage in human factor XI, explaining why rabbit factor XI is a monomer in non-reducing SDS/PAGE. On size-exclusion chromatography, however, purified plasma rabbit factor XI, like the human protein and unlike prekallikrein, eluted as a dimer, demonstrating that rabbit factor XI circulates as a non-covalent dimer. In functional assays rabbit factor XI and human factor XI behaved similarly. Both monomeric and dimeric factor XI were detected in extracts of cells expressing rabbit factor XI. We conclude that the failure of rabbit factor XI to form a covalent homodimer due to the replacement of Cys-321 with His does not impair its functional activity because it exists in plasma as a non-covalent homodimer and homodimerization is an intracellular process.

CDNA cloning and expression of human activin βE subunit

We cloned human activin βE subunit cDNA from a liver cDNA library using the polymerase chain reaction (PCR) technique. The deduced amino acid sequence was 97 and 96% homologous to the mouse and rat activin βE subunits. Human activin βE subunit tagged with Myc and polyhistidine residues at the COOH terminus was expressed in mammalian cells and secreted into the medium as a disulphide-linked homodimer protein. We also found that the human activin βE protein could bind to follistatin, an activin-binding protein. Northern blot analysis showed that this gene was expressed as a major transcript of 2.7 kb predominantly in human liver. These findings suggest that activin E (dimeric protein) may play a role in humans.

PDGF-D is a specific, protease-activated ligand for the PDGF β-receptor

The term 'platelet-derived growth factor' (PDGF) refers to a family of disulphide-bonded dimeric isoforms that are important for growth, survival and function in several types of connective tissue cell. So far, three different PDGF chains have been identified - the classical PDGF-A and PDGF-B and the recently identified PDGF-C. PDGF isoforms (PDGF-AA, AB, BB and CC) exert their cellular effects by differential binding to two receptor tyrosine kinases. The PDGF alpha-receptor (PDGFR-alpha) binds to all three PDGF chains, whereas the beta-receptor (PDGFR-beta) binds only to PDGF-B. Gene-targeting studies using mice have shown that the genes for PDGF-A and PDGF-B, as well as the two PDGFR genes, are essential for normal development. Furthermore, overexpression of PDGFs is linked to different pathological conditions, including malignancies, atherosclerosis and fibroproliferative diseases. Here we have identify and characterize a fourth member of the PDGF family, PDGF-D. PDGF-D has a two-domain structure similar to PDGF-C and is secreted as a disulphide-linked homodimer, PDGF-DD. Upon limited proteolysis, PDGF-DD is activated and becomes a specific agonistic ligand for PDGFR-beta. PDGF-DD is the first known PDGFR-beta-specific ligand, and its unique receptor specificity indicates that it may be important for development and pathophysiology in several organs.

The solution structure of the disulphide-linked homodimer of the human trefoil protein TFF1

The trefoil factor family protein, TFF1, forms a homodimer, via a disulphide linkage, that has greater activity in wound healing assays than the monomer. Having previously determined a high-resolution solution structure of a monomeric analogue of TFF1, we now investigate the structure of the homodimer formed by the native sequence. The two putative receptor/ligand recognition domains are found to be well separated, at opposite ends of a flexible linker. This contrasts sharply with the known fixed and compact arrangement of the two trefoil domains of the closely related TFF2, and has significant implications for the mechanism of action and functional specificity of the TFF of proteins.

Structural basis for co-stimulation by the human CTLA-4/B7-2 complex.

Regulation of T-cell activity is dependent on antigen-independent co-stimulatory signals provided by the disulphide-linked homodimeric T-cell surface receptors, CD28 and CTLA-4 (ref. 1). Engagement of CD28 with B7-1 and B7-2 ligands on antigen-presenting cells (APCs) provides a stimulatory signal for T-cell activation, whereas subsequent engagement of CTLA-4 with these same ligands results in attenuation of the response. Given their central function in immune modulation, CTLA-4- and CD28-associated signalling pathways are primary therapeutic targets for preventing autoimmune disease, graft versus host disease, graft rejection and promoting tumour immunity. However, little is known about the cell-surface organization of these receptor/ligand complexes and the structural basis for signal transduction. Here we report the 3.2-A resolution structure of the complex between the disulphide-linked homodimer of human CTLA-4 and the receptor-binding domain of human B7-2. The unusual dimerization properties of both CTLA-4 and B7-2 place their respective ligand-binding sites distal to the dimer interface in each molecule and promote the formation of an alternating arrangement of bivalent CTLA-4 and B7-2 dimers that extends throughout the crystal. Direct observation of this CTLA-4/B7-2 network provides a model for the periodic organization of these molecules within the immunological synapse and suggests a distinct mechanism for signalling by dimeric cell-surface receptors.

Structural basis of glutamate recognition by a dimeric metabotropic glutamate receptor

The metabotropic glutamate receptors (mGluRs) are key receptors in the modulation of excitatory synaptic transmission in the central nervous system. Here we have determined three different crystal structures of the extracellular ligand-binding region of mGluR1--in a complex with glutamate and in two unliganded forms. They all showed disulphide-linked homodimers, whose 'active' and 'resting' conformations are modulated through the dimeric interface by a packed alpha-helical structure. The bi-lobed protomer architectures flexibly change their domain arrangements to form an 'open' or 'closed' conformation. The structures imply that glutamate binding stabilizes both the 'active' dimer and the 'closed' protomer in dynamic equilibrium. Movements of the four domains in the dimer are likely to affect the separation of the transmembrane and intracellular regions, and thereby activate the receptor. This scheme in the initial receptor activation could be applied generally to G-protein-coupled neurotransmitter receptors that possess extracellular ligand-binding sites.