Dimer-dimer Contacts Research Articles

Crystal structure of a bovine neurophysin II dipeptide complex at 2.8 A determined from the single-wavelength anomalous scattering signal of an incorporated iodine atom.

The crystal structure of a dipeptide complex of bovine neurophysin II has been solved at 2.8 A resolution solely by using single-wavelength anomalous scattering data from a single iodinated derivative. The asymmetric unit is an elongated tetramer of dimensions 110 x 40 x 30 A, composed of two dimers related by pseudo twofold symmetry. Each monomer consists of two homologous layers, each with four antiparallel beta-strands. The two regions are connected by a helix followed by a long loop. Monomer-monomer contacts involve antiparallel beta-sheet interactions, which form a dimer with two layers of eight beta-strands. One peptide per monomer occupies the principal hormone-binding pocket formed by part of the amino-terminal region and parts of the connecting helix and loop, with binding to protein consistent with conclusions drawn from solution studies. Dimer-dimer contacts involve the Tyr49 region adjacent to this site. A fifth dipeptide, of unknown biological significance, helps to stabilize one of the monomer-monomer interfaces and the tetramer-tetramer network in the crystal.

The Ff gene 5 protein-d(pA)40-60 complex: 1H NMR supports a localized base-binding model.

The interaction between Ff gene 5 protein (G5P) and d(pA)40-60 serves as an improved model system for a 1H NMR examination of the G5P-ssDNA interface under cooperative binding conditions. Selective deuteriation of aromatic residues enables individual Tyr (3,5)H and (2,6)H resonances to be monitored in spectra of high molecular weight nucleoprotein assemblies. Analysis of complexation-induced chemical shift changes and intermolecular NOEs indicates that Tyr 26 is the only tyrosine to interact directly with ssDNA. Tyr 41, which is immobilized upon binding, is implicated in a dimer-dimer contact role. These and other NMR data are consistent with a previously outlined model of the protein-DNA interface in which Phe 73', Leu 28, and Tyr 26 form components of a base-binding pocket or "dynamic clamp" fringed by a cluster of positively charged residues [King, G. C., & Coleman, J. E. (1987) Biochemistry 26, 2929-2937]. In the present version of this model, the Phe and Leu side chains are proposed to stack on either side of a single base, while there is the possibility that Tyr 26 may H-bond to the sugar-phosphate backbone in addition to or instead of stacking. Chemical-exchange effects underscore the dynamic nature of binding at the pocket. A comparison of d(pA)40-60 and oligo(dA)-induced chemical shift changes suggests that poly- and oligonucleotide complexes have indistinguishable base-binding loci but appear to differ in their dimer-dimer interactions.(ABSTRACT TRUNCATED AT 250 WORDS)

A model for intracellular complexation between gene-5 protein and bacteriophage fd DNA.

A structural model for the helical intracellular complex formed between the gene-5 DNA-binding protein (G 5 BP; approximately 1274 copies) and bacteriophage fd DNA has been derived by an atomic-contact analysis approach. These studies depended in large part on the recently determined high-resolution structure of the G 5 BP dimer and cross-correlations with physical-chemical data available from other techniques. The approach was to systematically scan the full set of helical complexation parameters involved, based upon observed structural and orientational constraints, to determine those compatible with both the structure of the G 5 BP dimer and the overall dimensions of the full complex. This process was monitored throughout by close scrutiny of dimer-dimer contacts and the use of hard-copy and interactive graphics devices. Instead of the wide variety of possibilities that had been expected from such an approach, only one satisfactory assembly of DNA and G 5 BP dimers could be found. The results indicate that phage DNA will be wound to the outside of the helical protein ribbon that forms the core of intracellular complex at a density of five nucleotides per G 5 BP monomer. Bound DNA strands are positioned in two contiguous binding channels, which form as a consequence of the interactions of complexed G 5 BP dimers. These channels run just inside the outer extended beta loops, composed of residue 20-30, and are separated by approximately 3.2 nm. The DNA phosphate backbone is bound at a substantially smaller radial distance (approximately 3.5 nm) than the maximum radius of the intracellular complex as a whole (approximately 4.5 nm) since bound DNA is embedded within these well-defined binding channels. Our studies also indicate that a number of sterically unacceptable contacts, involving residues 38-42, prevent complexation of otherwise complementary dimer surfaces in the absence of nucleic acids. In the process of binding DNA, these residues change conformation thereby allowing self-assembly of dimer units into a helical structure. We propose that these residues act as a two-position stereochemical switch that allows or disallows complex formation in response to the absence or presence of DNA.

Molecular Heterogeneity of Hereditary Pyropoikilocytosis: Identification of a Second Variant of the Spectrin α-Subunit

In hereditary pyropoikilocytosis (HPP), the red cell membrane skeletons exhibit a mechanical instability that can be correlated to defective self-association of spectrin heterodimers. To determine the underlying molecular defect, we have subjected HPP spectrin to limited tryptic digestion, followed by one- and two-dimensional separations of the peptides. Two of the HPP kindreds exhibited a marked decrease in 80,000-dalton peptide (previously identified as the spectrin dimer-dimer contact domain of the alpha-subunit) and a concomitant increase of the 74,000-dalton polypeptide (presumably derived from the 80,000-dalton domain) and a decrease in a 22,000-dalton polypeptide. We now report tryptic digests of two other HPP kindred that are characterized by a decrease or complete absence of the 80,000-dalton tryptic fragment, with a concomitant increase in fragments at 46,000 and 17,000 daltons. The 46,000-dalton fragment separated into multiple spots on isoelectric focusing, ranging in isoelectric point from 5.25 to 5.35, and the 17,000-dalton fragment focused to a single spot at 5.4. Minor fragments at 56,000 and 22,000 daltons were also decreased, while a 38,000-dalton fragment increased. Limited tryptic digestion of the separated alpha- and beta-subunits revealed that the 74,000-dalton fragment in the first group of patients and the 46,000-dalton fragment in the second group of patients were derived from the alpha-subunit. Both subtypes exhibited a similar defect of spectrin self-association, with 30%-38% of spectrin dimers in O degrees C extracts. The results indicate that at least two distinct forms of structurally defective spectrin may give rise to the clinical presentation of HPP.

Molecular heterogeneity of hereditary pyropoikilocytosis: identification of a second variant of the spectrin alpha-subunit

In hereditary pyropoikilocytosis (HPP), the red cell membrane skeletons exhibit a mechanical instability that can be correlated to defective self-association of spectrin heterodimers. To determine the underlying molecular defect, we have subjected HPP spectrin to limited tryptic digestion, followed by one- and two-dimensional separations of the peptides. Two of the HPP kindreds exhibited a marked decrease in 80,000- dalton peptide (previously identified as the spectrin dimer-dimer contact domain of the alpha-subunit) and a concomitant increase of the 74,000-dalton polypeptide (presumably derived from the 80,000-dalton domain) and a decrease in a 22,000-dalton polypeptide. We now report tryptic digests of two other HPP kindred that are characterized by a decrease or complete absence of the 80,000-dalton tryptic fragment, with a concomitant increase in fragments at 46,000 and 17,000 daltons. The 46,000-dalton fragment separated into multiple spots on isoelectric focusing, ranging in isoelectric point from 5.25 to 5.35, and the 17,000-dalton fragment focused to a single spot at 5.4. Minor fragments at 56,000 and 22,000 daltons were also decreased, while a 38,000-dalton fragment increased. Limited tryptic digestion of the separated alpha- and beta-subunits revealed that the 74,000-dalton fragment in the first group of patients and the 46,000-dalton fragment in the second group of patients were derived from the alpha-subunit. Both subtypes exhibited a similar defect of spectrin self-association, with 30%-38% of spectrin dimers in O degrees C extracts. The results indicate that at least two distinct forms of structurally defective spectrin may give rise to the clinical presentation of HPP.

The interaction of insulin with its receptor: cross-linking via insulin association as the source of receptor clustering.

The extensive association of mammalian insulins in solution and the aggregation of insulin receptors in cell membranes are well documented. The hypothesis advanced here is that a direct connection exists between these observations. It is postulated that, after binding to its receptor, an insulin monomer can interact with another similarly bonded hormone-receptor complex through those groups on the insulin monomer faces utilized for dimer-dimer contacts in the crystal and in solution. Regarded thus, the insulin molecules are effectively bivalent as required for the formation of cross-links between receptors, with the accompanying enhancement of biological activity. A number of properties of native insulins from different animals, and of modified insulins, are considered in the light of this suggestion. It is shown to have considerable power in reconciling a diversity of such observations and to provide a plausible model for the experimentally observed receptor clustering phenomenon.

Thermodynamic studies on subunit assembly in human hemoglobin. Temperature dependence of the dimer-tetramer association constants for oxygenated and unliganded hemoglobins.

Equilibrium constants for dimer-tetramer association of oxygenated human hemoglobin have been determined as a function of temperature, by analytical molecular sieve chromatography, yielding a van't Hoff enthalpy of 3.8 +/- 1.6 kcal/mol, and a unitary entropy of 48.4 +/- 5.7 e.u. Under the same conditions (0.1 M Tris/HCl, 0.1 M NaCl, 1 mM Na2EDTA, pH 7.4) the equilibrium constants for dimer-tetramer association of unliganded hemoglobin have been determined as a function of temperature from kinetic studies of the forward and reverse rate constants. The derived enthalpy and unitary entropy for association of unliganded hemoglobin are 0deltaH2 = 28.9 +/- 0.5 kcal/mol and 0deltaS2 = -41.8 +/- 1.7 e.u., respectively. Thus the oxygenation-linked enthalpy of subunit association is 32.7 +/- 1.7 kcal. The corresponding entropic coupling is 90.2 +/- 5.9 e.u. The thermodynamic parameters are consistent with an increased role of hydrophobic interactions within the dimer-dimer contact region upon oxygenation, or a decreased role of hydrogen bonds and ion pair interactions. The opposite signs for association entropies in unliganded and oxygenated states suggest the existence of two distinct subunit interaction processes during the course of oxygenation with a reversal in sign at some particular binding step. The results provide a necessary basis for interpretation of calorimetric results on reconstitution of hemoglobin, described in an accompanying paper.

Kinetics of deoxyhemoglobin subunit dissociation determined by haptoglobin binding: estimation of the equilibrium constant from forward and reverse rates.

Deoxyhemoglobin tetramers dissociate into dimers very slowly, with half-times on the order of several hours. It is demonstrated that absorbance changes in the Soret region which accompany this dissociation and persist upon binding of haptoglobin 1-1 to the dissociated dimers can be used for accurate kinetic determinations over the necessarily long periods required for study. This method of study for the slow reactions depends upon long-term spectral integrity of the reaction mixtures and upon accurate measurement. The variation in rate constants determined by this procedure has been correlated with variations in structural constraints at the dimer-dimer contact region. In the presence of 2,3-diphosphoglycerate the rate constant is decreased, consistent with the role of this effector in binding to both beta chains and stabilizing the constrained deoxy tetramer against dissociation into alphabeta dimers. With hemoglobin specifically modified (des-Arg-141alpha) to eliminate half the constraining salt links within the dimer-dimer contact region, the dissociation rate is increased by approximately three orders of magnitude. In hemoglobin S where the amino acid substitution is not directly in the intersubunit contact region of interest, the dissociation rate is found to be approximately the same as that for hemoglobin A. Combination of the dissociation rate constants determined by haptoglobin binding with stopped-flow determinations of the rate constant for reassociation of dissociated dimers provides an estimate of the equilibrium constant, 0K2, for the deoxyhemoglobin dimer-tetramer equilibrium. This estimate is independent of any assumptions regarding other energetic quantities, and yields a value of 2.54 +/- 0.7 X 10(10)M-1 (heme) in 0.1 M Tris-HCl, 0.1 M NaCl, and 1 mM EDTA, pH 7.4, 21.5 degrees C. Thus the intersubunit contact energy is -14.0 +/- 0.2 kcal/mol of heme. The stabilization energy between deoxy and oxy tetramers is found to be approximately 6.4 kcal/mol, under these conditions.

The linkage between oxygenation and subunit dissociation in human hemoglobin.

The use of subunit dissociation as a means of probing intersubunit contact energy changes which accompany cooperative ligand binding has been studied for the case of human hemoglobin. An analysis is presented delineating the information that can be obtained from the linkage relationships between ligand binding and subunit dissociation of hemoglobin tetramers into dimers. The analysis defines (a) the variation of the saturation function, Y, with total protein concentration, (b) the variation of the subunit dissociation constant (x)K(2) with ligand concentration (X) and (c) the correlations between changes in dimer-dimer contact energy and the sequential ligand binding steps. Sensitivity of the linkage function has been explored by numerical simulation. It is shown that subunit dissociation may appreciably affect oxygenation curves under usual conditions of measurement and that relying solely on either (x)K(2) or Y may lead to incorrect picutres of the energetics, whereas the combination defines the system much more exactly.