Retinoid X receptor alpha forms tetramers in solution.

Protein-protein interactions allow the retinoid X receptor (RXR) to bind to cognate DNA as a homo- or a heterodimer and to participate in mediating the effects of a variety of hormones on gene transcription. Here we report a systematic study of the oligomeric state of RXR in the absence of a DNA template. We have used electrophoresis under nondenaturing conditions and chemical crosslinking to show that in solution, RXR alpha forms homodimers as well as homotetramers. The dissociation constants governing dimer and tetramer formation were estimated by fluorescence anisotropy studies. The results indicate that RXR tetramers are formed with a high affinity and that at protein concentrations higher than about 70 nM, tetramers will constitute the predominant species. Tetramer formation may provide an additional level of the regulation of gene transcription mediated by RXRs.

A variant of lambda repressor with an altered pattern of cooperative binding to DNA sites.

The bacteriophage lambda repressor binds cooperatively to pairs of adjacent sites in the lambda chromosome, one repressor dimer binding to each site. The repressor's amino domain (that which mediates DNA binding) is connected to its carboxyl domain (that which mediates dimerization and the interaction between dimers) by a protease-sensitive linker region. We have generated a variant lambda repressor that lacks this linker region. We show that dimers of the variant protein are deficient in cooperative binding to sites at certain, but not all, distances. The linker region thus extends the range over which carboxyl domains of DNA-bound dimers can interact. In particular, the linker is required for cooperative binding to a pair of sites as found in the lambda chromosome, and thus is essential for the repressor's physiological function.

Repacking protein cores with backbone freedom: structure prediction for coiled coils.

Progress in homology modeling and protein design has generated considerable interest in methods for predicting side-chain packing in the hydrophobic cores of proteins. Present techniques are not practically useful, however, because they are unable to model protein main-chain flexibility. Parameterization of backbone motions may represent a general and efficient method to incorporate backbone relaxation into such fixed main-chain models. To test this notion, we introduce a method for treating explicitly the backbone motions of alpha-helical bundles based on an algebraic parameterization proposed by Francis Crick in 1953 [Crick, F. H. C. (1953) Acta Crystallogr. 6, 685-689]. Given only the core amino acid sequence, a simple calculation can rapidly reproduce the crystallographic main-chain and core side-chain structures of three coiled coils (one dimer, one trimer, and one tetramer) to within 0.6-A root-mean-square deviations. The speed of the predictive method [approximately 3 min per rotamer choice on a Silicon Graphics (Mountain View, CA) 4D/35 computer] permits it to be used as a design tool.

The CuA center of cytochrome-c oxidase: electronic structure and spectra of models compared to the properties of CuA domains.

The electronic structure and spectrum of several models of the binuclear metal site in soluble CuA domains of cytochrome-c oxidase have been calculated by the use of an extended version of the complete neglect of differential overlap/spectroscopic method. The experimental spectra have two strong transitions of nearly equal intensity around 500 nm and a near-IR transition close to 800 nm. The model that best reproduces these features consists of a dimer of two blue (type 1) copper centers, in which each Cu atom replaces the missing imidazole on the other Cu atom. Thus, both Cu atoms have one cysteine sulfur atom and one imidazole nitrogen atom as ligands, and there are no bridging ligands but a direct Cu-Cu bond. According to the calculations, the two strong bands in the visible region originate from exciton coupling of the dipoles of the two copper monomers, and the near-IR band is a charge-transfer transition between the two Cu atoms. The known amino acid sequence has been used to construct a molecular model of the CuA site by the use of a template and energy minimization. In this model, the two ligand cysteine residues are in one turn of an alpha-helix, whereas one ligand histidine is in a loop following this helix and the other one is in a beta-strand.

One sequence, two folds: a metastable structure of CD2.

When expressed as part of a glutathione S-transferase fusion protein the NH2-terminal domain of the lymphocyte cell adhesion molecule CD2 is shown to adopt two different folds. The immunoglobulin superfamily structure of the major (85%) monomeric component has previously been determined by both x-ray crystallography and NMR spectroscopy. We now describe the structure of a second, dimeric, form present in about 15% of recombinant CD2 molecules. After denaturation and refolding in the absence of the fusion partner, dimeric CD2 is converted to monomer, illustrating that the dimeric form represents a metastable folded state. The crystal structure of this dimeric form, refined to 2.0-A resolution, reveals two domains with overall similarity to the IgSF fold found in the monomer. However, in the dimer each domain is formed by the intercalation of two polypeptide chains. Hence each domain represents a distinct folding unit that can assemble in two different ways. In the dimer the two domains fold around a hydrophilic interface believed to mimic the cell adhesion interaction at the cell surface, and the formation of dimer can be regulated by mutating single residues at this interface. This unusual misfolded form of the protein, which appears to result from inter- rather than intramolecular interactions being favored by an intermediate structure formed during the folding process, illustrates that evolution of protein oligomers is possible from the sequence for a single protein domain.

Identification of four acidic amino acids that constitute the catalytic center of the RuvC Holliday junction resolvase.

Escherichia coli RuvC protein is a specific endonuclease that resolves Holliday junctions during homologous recombination. Since the endonucleolytic activity of RuvC requires a divalent cation and since 3 or 4 acidic residues constitute the catalytic centers of several nucleases that require a divalent cation for the catalytic activity, we examined whether any of the acidic residues of RuvC were required for the nucleolytic activity. By site-directed mutagenesis, we constructed a series of ruvC mutant genes with similar amino acid replacements in 1 of the 13 acidic residues. Among them, the mutant genes with an alteration at Asp-7, Glu-66, Asp-138, or Asp-141 could not complement UV sensitivity of a ruvC deletion strain, and the multicopy mutant genes showed a dominant negative phenotype when introduced into a wild-type strain. The products of these mutant genes were purified and their biochemical properties were studied. All of them retained the ability to form a dimer and to bind specifically to a synthetic Holliday junction. However, they showed no, or extremely reduced, endonuclease activity specific for the junction. These 4 acidic residues, which are dispersed in the primary sequence, are located in close proximity at the bottom of the putative DNA binding cleft in the three-dimensional structure. From these results, we propose that these 4 acidic residues constitute the catalytic center for the Holliday junction resolvase and that some of them play a role in coordinating a divalent metal ion in the active center.

An antiestrogen: a phosphotyrosyl peptide that blocks dimerization of the human estrogen receptor.

We have previously identified tyrosine-537 as a constitutively phosphorylated site on the human estrogen receptor (hER). A 12-amino acid phosphotyrosyl peptide containing a selected sequence surrounding tyrosine-537 was used to investigate the function of phosphotyrosine-537. The phosphotyrosyl peptide completely blocked the binding of the hER to an estrogen response element (ERE) in a gel mobility shift assay. Neither the nonphosphorylated tyrosyl peptide nor an unrelated phosphotyrosyl peptide previously shown to inhibit the signal transducers and activators of transcription factor (STAT) blocked binding of the hER to the ERE. The hER phosphotyrosyl peptide was shown by molecular sizing chromatography to dissociate the hER dimer into monomers. The hER specifically bound the 32P-labeled phosphotyrosyl peptide, indicating that the inhibition of ERE binding was caused by the phosphotyrosyl peptide binding directly to the hER and blocking dimerization. These data suggest that the phosphorylation of tyrosine-537 is a necessary step for the formation of the hER dimer. In addition, we propose that the dimerization of the hER occurs by a previously unrecognized Src homology 2 domain (SH2)-like phosphotyrosyl coupling mechanism. Consequently, the phosphotyrosyl peptide represents a class of antagonists that inhibits estrogen action by a mechanism other than interacting with the receptor's hormone binding site.

Transition metal ion activation of DNA binding by the diphtheria tox repressor requires the formation of stable homodimers.

The diphtheria tox repressor (DtxR) is a transition metal ion-dependent regulatory element that controls the expression of diphtheria toxin and several genes involved in the synthesis of siderophores in Corynebacterium diphtheriae. In the presence of transition metal ions apo-DtxR becomes activated and specifically binds to its target DNA sequences. We demonstrate by glutaraldehyde cross-linking that monomeric apo-DtxR is in weak equilibrium with a dimeric form and that upon addition of activating metal ions to the reaction mixture a dimeric complex is stabilized. Addition of the DNA-binding-defective mutant apo-DtxR(delta 1-47) to apo-DtxR in the absence of transition metal ions inhibits conversion of the apo-repressor to its activated DNA-binding form. We also show that the binding of Ni2+ to both apo-DtxR and apo-DtxR(delta 1-47) is cooperative and that upon ion binding there is a conformational change in the environment of the indole ring moiety of Trp-104. For the wild-type repressor the consequences of this conformational change include a shift in equilibrium toward dimer formation and activation of target DNA binding by the repressor. We conclude that the formation of DtxR homodimers is mediated through a protein-protein interaction domain that is also activated on metal ion binding.

Function of the p86 subunit of eukaryotic initiation factor (iso)4F as a microtubule-associated protein in plant cells.

The isozyme form of eukaryotic initiation factor 4F [eIF-(iso)4F] from wheat germ is composed of a p28 subunit that binds the 7-methylguanine cap of mRNA and a p86 subunit having unknown function. The p86 subunit was found to have limited sequence similarity to a kinesin-like protein encoded by the katA gene of Arabidopsis thaliana. Native wheat germ eIF-(iso)4F and bacterially expressed p86 subunit and p86-p28 complex bound to taxol-stabilized maize microtubules (MTs) in vitro. Binding saturation occurred at 1 mol of p86 per 5-6 mol of polymerized tubulin dimer, demonstrating a substoichiometric interaction of p86 with MTs. No evidence was found for a direct interaction of the p28 subunit with MTs. Unlike kinesin, cosedimentation of eIF-(iso)4F with MTs was neither reduced by MgATP nor enhanced by adenosine 5'-[gamma-imido]triphosphate. Both p86 subunit and p86-p28 complex induced the bundling of MTs in vitro. The p86 subunit was immunolocalized to the cytosol in root maize cells and existed in three forms: fine particles, coarse particles, and linear patches. Many coarse particles and linear patches were colocalized or closely associated with cortical MT bundles in interphase cells. The results indicate that the p86 subunit of eIF-(iso)4F is a MT-associated protein that may simultaneously link the translational machinery to the cytoskeleton and regulate MT disposition in plant cells.

ncd and kinesin motor domains interact with both alpha- and beta-tubulin.

Motor domains of the Drosophila minus-end-directed microtubule (MT) motor protein ncd, were found to saturate microtubule binding sites at a stoichiometry of approximately one motor domain per tubulin dimer. To determine the tubulin subunit(s) involved in binding to ncd, mixtures of ncd motor domain and MTs were treated with the zero-length cross-linker 1-ethyl-3-(3-dimethylaminopropyl-carbodiimide) (EDC). EDC treatment generated covalently cross-linked products of ncd and alpha-tubulin and of ncd and beta-tubulin, indicating that the ncd motor domain interacts with both alpha- and beta-tubulin. When the Drosophila kinesin motor domain protein was substituted for the ncd motor domain, cross-linked products of kinesin and alpha-tubulin and of kinesin and beta-tubulin were produced. EDC treatment of mixtures of ncd motor domain and unassembled tubulin dimers or of kinesin motor domain and unassembled tubulin dimers produced the same motor-tubulin products generated in the presence of MTs. These results indicate that kinesin family motors of opposite polarity interact with both tubulin monomers and support a model in which some portion of each protein's motor domain overlaps adjacent alpha- and beta-tubulin subunits.

Chimeric tumor necrosis factor receptors with constitutive signaling activity.

Many hormone and cytokine receptors are crosslinked by their specific ligands, and multimerization is an essential step leading to the generation of a signal. In the case of the tumor necrosis factor (TNF) receptors (TNF-Rs), antibody-induced crosslinking is sufficient to trigger a cytolytic effect. However, the quaternary structural requirements for signaling--i.e., the formation of dimers, trimers, or higher-order multimers--have remained obscure. Moreover, it has not been clear whether the 55-kDa or 75-kDa TNF-R is responsible for initiation of cytolysis. We reasoned that an obligate receptor dimer, targeted to the plasma membrane, might continuously signal the presence of TNF despite the actual absence of the ligand. Such a molecule, inserted into an appropriate vector, could be used to project receptor-specific "TNF-like" activity to specific cells and tissues in vivo. Accordingly, we constructed sequences encoding chimeric receptors in which the extracellular domain of the mouse erythropoietin receptor (Epo-R) was fused to the "stem," transmembrane domain, and cytoplasmic domain of the two mouse TNF-Rs. Thus, the Epo-R group was used to drive dimerization of the TNF-R cytoplasmic domain. These chimeric proteins were well expressed in a variety of cell lines and bound erythropoietin at the cell surface. Both the 55-kDa and the 75-kDa Epo/TNF-R chimeras exerted a constitutive cytotoxic effect detected by cotransfection or clonogenic assay. Thus, despite the lack of structural homology between the cytoplasmic domains of the two TNF-Rs, a similar signaling endpoint was observed. Moreover, dimerization (rather than trimerization or higher-order multimerization) was sufficient for elicitation of a biological response.

Complex flexibility of the transforming growth factor beta superfamily.

The transforming growth factors beta (TGF-beta s) are important modulators of growth and differentiation. They are intermolecular disulfide-bonded homodimeric molecules. The monomer fold has a conserved cystine knot and lacks a hydrophobic core. The biological specificity of a given member of the family is believed to be determined by the conformational flexibility of the variable loop regions of the monomer. The monomer subunit assembly in the dimer is stabilized mainly by hydrophobic contacts and a few hydrogen bonds. Since these interactions are nondirectional, we examined subunit assemblies of TGF-beta by using conformational analysis. The different subunit assemblies in TGF-beta 2 dimer were characterized in terms of the intersubunit disulfide torsion. Our analyses show that the subunit assemblies fall into two states: the crystallographically observed gauche+conformation and the previously not reported gauche--conformation, both having almost identical interaction energies. Furthermore, there is significant flexibility in the subunit assembly within the gauche+ and the gauche- states of the disulfide bond. The monomer subunit assembly is independent of the variations about the loop regions. The variations in the loop regions, coupled with flexibility in the monomer assembly, lead to a complex flexibility in the dimer of the TGF-beta superfamily. For the TGF-beta superfamily, the cystine knot acts as a scaffold and complex flexibility provides for biological selectivity. Complex flexibility might provide an explanation for the diverse range of biological activities that these important molecules display.

A multimer model for P680, the primary electron donor of photosystem II.

We consider a model of the photosystem II (PS II) reaction center in which its spectral properties result from weak (approximately 100 cm-1) excitonic interactions between the majority of reaction center chlorins. Such a model is consistent with a structure similar to that of the reaction center of purple bacteria but with a reduced coupling of the chlorophyll special pair. We find that this model is consistent with many experimental studies of PS II. The similarity in magnitude of the exciton coupling and energetic disorder in PS II results in the exciton states being structurally highly heterogeneous. This model suggests that P680, the primary electron donor of PS II, should not be considered a dimer but a multimer of several weakly coupled pigments, including the pheophytin electron acceptor. We thus conclude that even if the reaction center of PS II is structurally similar to that of purple bacteria, its spectroscopy and primary photochemistry may be very different.

Swapping structural determinants of ribonucleases: an energetic analysis of the hinge peptide 16-22.

Bovine seminal ribonuclease (BS-RNase) is a homodimeric enzyme strictly homologous to the pancreatic ribonuclease (RNase A). Native BS-RNase is an equilibrium mixture of two distinct dimers differing in the interchange of the N-terminal segments and in their biological properties. The loop 16-22 plays a fundamental role on the relative stability of the two isomers. Both the primary and tertiary structures of the RNase A differ substantially from those of the seminal ribonuclease in the loop region 16-22. To analyze the possible stable conformations of this loop in both enzymes, structure predictions have been attempted, according to a procedure described by Palmer and Scheraga [Palmer, K. A. & Scheraga, H. A. (1992) J. Comput. Chem. 13, 329-350]. Results compare well with experimental x-ray structures and clarify the structural determinants that are responsible for the swapping of the N-terminal domains and for the peculiar properties of BS-RNase. Minimal modifications of RNase A sequence needed to form a stable swapped dimer are also predicted.

Intersubunit contacts made by tryptophan 120 with biotin are essential for both strong biotin binding and biotin-induced tighter subunit association of streptavidin.

In natural streptavidin, tryptophan 120 of each subunit makes contacts with the biotin bound by an adjacent subunit through the dimer-dimer interface. To understand quantitatively the role of tryptophan 120 and its intersubunit communication in the properties of streptavidin, a streptavidin mutant in which tryptophan 120 is converted to phenylalanine was produced and characterized. The streptavidin mutant forms a tetrameric molecule and binds one biotin per subunit, as does natural streptavidin, indicating that the mutation of tryptophan 120 to phenylalanine has no significant effect on the basic properties of streptavidin. However, its biotin-binding affinity was reduced substantially, to approximately 10(8) M-1, indicating that the contact made by tryptophan 120 to biotin has a considerable contribution to the extremely tight biotin binding by streptavidin. The mutant retained bound biotin over a wide pH range or with the addition of urea up to 6 M at neutral pH. However, bound biotin was efficiently released by the addition of excess free biotin due, presumably, to exchange reactions. Electrophoretic analysis revealed that the intersubunit contact made by tryptophan 120 to biotin through the dimer-dimer interface is the major interaction responsible for the biotin-induced, tighter subunit association of streptavidin. In addition, the mutant has weaker subunit association than natural streptavidin even in the absence of biotin, indicating that tryptophan 120 also contributes to the subunit association of tetramers in the absence of biotin.