Insulin gene enhancer binding protein Isl-1 is a member of a novel class of proteins containing both a homeo- and a Cys-His domain.

The activity of the rat insulin I gene enhancer is mainly dependent on two cis-acting protein-binding domains. Here we report the isolation of a complementary DNA encoding a protein, Isl-1, that binds to one of these domains. Isl-1 contains a homeodomain with greatest similarity to those of the Caenorhabditis elegans proteins encoded by mec-3 and lin-11. In addition, Isl-1, like the lin-11 and mec-3 gene products, contains a novel Cys-His domain which is reminiscent of known metal-binding regions. Together these proteins define a novel class of proteins containing both a homeo- and a Cys His-domain. Isl-1 is preferentially expressed in cells of pancreatic endocrine origin. If the structural homologies between Isl-1 and the C. elegans gene products reflect functional similarities, a role for Isl-1 in the development of pancreatic endocrine cells could be envisaged.     

Spatial relationship of the Fis binding sites for Hin recombinational enhancer activity

Site-specific recombination reactions involve the joining or rearrangement of discrete DNA segments in a highly precise manner. A site-specific DNA inversion regulates the expression of flagellin genes in Salmonella by switching the orientation of a promoter. Analysis of the reaction has shown that, in addition to DNA sequences at the two boundaries of the 1-kilobase invertible segment where strand exchange occurs, another cis acting sequence is required for efficient inversion. This 60-base-pair enhancer-like sequence can function at many different locations and in either orientation in a plasmid substrate. It includes two binding sites for a host protein called Factor II or Fis (refs 4 and 5). Here we have investigated the importance of the spatial relationship between the two Fis binding sites for enhancer activity and have found that the correct helical positioning of the binding sites on the DNA is critical. However, this result could not be accounted for by effects on Fis binding. We propose a model for enhancer function in which the enhancer region acts to align the recombination sites into a specific conformation required for productive synapsis.     

Structural basis of RNA recognition and activation by innate immune receptor RIG-I

Retinoic-acid-inducible gene-I (RIG-I; also known as DDX58) is a cytoplasmic pathogen recognition receptor that recognizes pathogen-associated molecular pattern (PAMP) motifs to differentiate between viral and cellular RNAs. RIG-I is activated by blunt-ended double-stranded (ds)RNA with or without a 5'-triphosphate (ppp), by single-stranded RNA marked by a 5'-ppp and by polyuridine sequences. Upon binding to such PAMP motifs, RIG-I initiates a signalling cascade that induces innate immune defences and inflammatory cytokines to establish an antiviral state. The RIG-I pathway is highly regulated and aberrant signalling leads to apoptosis, altered cell differentiation, inflammation, autoimmune diseases and cancer. The helicase and repressor domains (RD) of RIG-I recognize dsRNA and 5'-ppp RNA to activate the two amino-terminal caspase recruitment domains (CARDs) for signalling. Here, to understand the synergy between the helicase and the RD for RNA binding, and the contribution of ATP hydrolysis to RIG-I activation, we determined the structure of human RIG-I helicase-RD in complex with dsRNA and an ATP analogue. The helicase-RD organizes into a ring around dsRNA, capping one end, while contacting both strands using previously uncharacterized motifs to recognize dsRNA. Small-angle X-ray scattering, limited proteolysis and differential scanning fluorimetry indicate that RIG-I is in an extended and flexible conformation that compacts upon binding RNA. These results provide a detailed view of the role of helicase in dsRNA recognition, the synergy between the RD and the helicase for RNA binding and the organization of full-length RIG-I bound to dsRNA, and provide evidence of a conformational change upon RNA binding. The RIG-I helicase-RD structure is consistent with dsRNA translocation without unwinding and cooperative binding to RNA. The structure yields unprecedented insight into innate immunity and has a broader impact on other areas of biology, including RNA interference and DNA repair, which utilize homologous helicase domains within DICER and FANCM.     

Coupling of agonist binding to channel gating in an ACh-binding protein linked to an ion channel

Neurotransmitter receptors from the Cys-loop superfamily couple the binding of agonist to the opening of an intrinsic ion pore in the final step in rapid synaptic transmission. Although atomic resolution structural data have recently emerged for individual binding and pore domains, how they are linked into a functional unit remains unknown. Here we identify structural requirements for functionally coupling the two domains by combining acetylcholine (ACh)-binding protein, whose structure was determined at atomic resolution, with the pore domain from the serotonin type-3A (5-HT3A) receptor. Only when amino-acid sequences of three loops in ACh-binding protein are changed to their 5-HT3A counterparts does ACh bind with low affinity characteristic of activatable receptors, and trigger opening of the ion pore. Thus functional coupling requires structural compatibility at the interface of the binding and pore domains. Structural modelling reveals a network of interacting loops between binding and pore domains that mediates this allosteric coupling process.     

Protein activity regulation by conformational entropy

How the interplay between protein structure and internal dynamics regulates protein function is poorly understood. Often, ligand binding, post-translational modifications and mutations modify protein activity in a manner that is not possible to rationalize solely on the basis of structural data. It is likely that changes in the internal motions of proteins have a major role in regulating protein activity, but the nature of their contributions remains elusive, especially in quantitative terms. Here we show that changes in conformational entropy can determine whether protein-ligand interactions will occur, even among protein complexes with identical binding interfaces. We have used NMR spectroscopy to determine the changes in structure and internal dynamics that are elicited by the binding of DNA to several variants of the catabolite activator protein (CAP) that differentially populate the inactive and active DNA-binding domain states. We found that the CAP variants have markedly different affinities for DNA, despite the CAP?DNA-binding interfaces being essentially identical in the various complexes. Combined with thermodynamic data, the results show that conformational entropy changes can inhibit the binding of CAP variants that are structurally poised for optimal DNA binding or can stimulate the binding activity of CAP variants that only transiently populate the DNA-binding-domain active state. Collectively, the data show how changes in fast internal dynamics (conformational entropy) and slow internal dynamics (energetically excited conformational states) can regulate binding activity in a way that cannot be predicted on the basis of the protein's ground-state structure.     

Requirement of hormone for thermal conversion of the glucocorticoid receptor to a DNA-binding state

A central question arising from the model of eukaryotic gene regulation by steroid hormone receptors is whether or not proteins represent pre-existing gene regulatory proteins that are activated on exposure to the extracellular signal. It has been generally believed that the ligand-binding of steroid hormone receptors triggers an allosteric change in receptor structure, manifested by an increased affinity of the receptor for DNA in vitro and nuclear target elements in vivo, as monitored by nuclear translocation. But this model has been challenged by recent reports indicating that glucocorticoid and progesterone receptors bind specifically in vitro to target DNA sequences even in the absence of hormone. On the other hand, it appears that the hormone induces protection in vivo of the glucocorticoid response element of the tyrosine amino transferase gene. Here we show that under conditions permitting minimal in vitro manipulation, the steroid-free glucocorticoid receptor in crude cytosol associates with the hsp90 heat shock protein (relative molecular mass Mr approximately equal to 90,000) to form a large 300K complex, rather than the 94K liganded receptor monomer. More importantly, we have developed an assay to demonstrate the requirement of hormone to dissociate the 300K complex by heat treatment. Specific DNA-binding activity of the receptor becomes apparent in this process, showing that DNA binding occurs but is inhibited in the large heteromeric complex. We propose a model in which receptor function is repressed by association of the receptor with hsp90. Dissociation of this complex is induced by the binding of steroid and is apparently an irreversible process.     

A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules

Class II and class I histocompatibility molecules allow T cells to recognize 'processed' polypeptide antigens. The two polypeptide chains of class II molecules, alpha and beta, are each composed of two domains (for review see ref. 6); the N-terminal domains of each, alpha 1 and beta 1, are highly polymorphic and appear responsible for binding peptides at what appears to be a single site and for being recognized by MHC-restricted antigen-specific T cells. Recently, the three-dimensional structure of the foreign antigen binding site of a class I histocompatibility antigen has been described. Because a crystal structure of a class II molecule is not available, we have sought evidence in class II molecules for the structural features observed in the class I binding site by comparing the patterns of conserved and polymorphic residues of twenty-six class I and fifty-four class II amino acid sequences. The hypothetical class II foreign-antigen binding site we present is consistent with mutation experiments and provides a structural framework for proposing peptide binding models to help understand recent peptide binding data.     

Peptide-induced conformational change of the class I heavy chain

There is evidence that peptide ligands take part in the assembly of class I molecules. In particular, addition of peptides to extracts of the mutant cells RMA-S and .174/T2, in which stable assembly of class I does not occur, results in a conformational change in the class I heavy chain and stable association of the heavy chain with beta 2-microglobulin (beta 2m). Thus specific peptides may stabilize or induce a conformational change in the class I heavy chain that results in a rise in the binding affinity of the heavy chain for beta 2m (Fig. 1a). Here we show that peptides have two cooperative roles in class I assembly. Specific short peptides (9-10 amino acids) can induce folding of the heavy chain in the absence of beta 2m. Both short (nine amino acids) and longer sequences (15 amino acids) can stabilize performed low-affinity complexes of heavy chain and beta 2m. To alter the conformation of free heavy chains, the peptides must be exactly the correct size, and they are found to correspond to the sequences isolated from infected cells. This property may therefore be the basis for selection of epitopes presented in vivo.     

Identification of human CD4 residues affecting class II MHC versus HIV-1 gp120 binding

Interactions of CD4 with the class II major histocompatibility complex (MHC) are crucial during thymic ontogeny and subsequently for helper and cytotoxic functions of CD4+CD8- T lymphocytes. CD4 is the receptor for the T-lymphotropic human immunodeficiency virus and binds its envelope glycoprotein, gp120. The residues involved in gp120 binding have been localized to a region within the immunoglobulin-like domain I of CD4, which corresponds to CDR2 of an immunoglobulin variable region, but the CD4 residues important in MHC class II interaction have not been characterized. Here, using a cell-binding assay dependent specifically on the CD4-MHC class II association, we analyse the effects of mutations in CD4 on class II versus gp120 binding. Mutations in CDR2 that destroy gp120 binding affect CD4-MHC class II binding similarly. In addition, binding of soluble gp120 to CD4-transfected cells abrogates their ability to interact with class II-bearing B lymphocytes. In contrast, other mutations within domains I or II that have no effect on gp120 binding eliminate or substantially decrease class II interaction. Thus, the CD4 binding site for class II MHC is more complex than the gp120 binding site, possibly reflecting a broader area of contact with the former ligand and a requirement for appropriate juxtaposition of the two N-terminal domains. The ability of gp120 to inhibit the binding of class II MHC to CD4 could be important in disrupting normal T-cell physiology, acting both to inhibit immune responses and to prevent differentiation of CD4+CD8+ thymocytes into CD4+CD8- T lymphocytes.     

A robust DNA mechanical device controlled by hybridization topology.

Controlled mechanical movement in molecular-scale devices has been realized in a variety of systems-catenanes and rotaxanes, chiroptical molecular switches, molecular ratchets and DNA-by exploiting conformational changes triggered by changes in redox potential or temperature, reversible binding of small molecules or ions, or irradiation. The incorporation of such devices into arrays could in principle lead to complex structural states suitable for nanorobotic applications, provided that individual devices can be addressed separately. But because the triggers commonly used tend to act equally on all the devices that are present, they will need to be localized very tightly. This could be readily achieved with devices that are controlled individually by separate and device-specific reagents. A trigger mechanism that allows such specific control is the reversible binding of DNA strands, thereby 'fuelling' conformational changes in a DNA machine. Here we improve upon the initial prototype system that uses this mechanism but generates by-products, by demonstrating a robust sequence-dependent rotary DNA device operating in a four-step cycle. We show that DNA strands control and fuel our device cycle by inducing the interconversion between two robust topological motifs, paranemic crossover (PX) DNA and its topoisomer JX2 DNA, in which one strand end is rotated relative to the other by 180 degrees. We expect that a wide range of analogous yet distinct rotary devices can be created by changing the control strands and the device sequences to which they bind.