Light scattering.

Light scattering methods can provide information about the native molecular weight, oligomeric composition, and gross conformation of a protein in solution. These methods are particularly well suited for studying large oligomeric systems or glycoproteins and can be used to characterize much larger structures involving protein such as viruses and even bacterial spores. This overview discusses theory and application of both static and dynamic light scattering analysis of protein solutions.

Formation and biochemical characterization of tube/pelle death domain complexes: critical regulators of postreceptor signaling by the Drosophila toll receptor.

In Drosophila, the Toll receptor signaling pathway is required for embryonic dorso-ventral patterning and at later developmental stages for innate immune responses. It is thought that dimerization of the receptor by binding of the ligand spätzle causes the formation of a postreceptor activation complex at the cytoplasmic surface of the membrane. Two components of this complex are the adaptor tube and protein kinase pelle. These proteins both have "death domains", protein interaction motifs found in a number of signaling pathways, particularly those involved in apoptotic cell death. It is thought that pelle is bound by tube during formation of the activation complexes, and that this interaction is mediated by the death domains. In this paper, we show using the yeast two-hybrid system that the wild-type tube and pelle death domains bind together. Mutant tube proteins which do not support signaling in the embryo are also unable to bind pelle in the 2-hybrid assay. We have purified proteins corresponding to the death domains of tube and pelle and show that these form corresponding heterodimeric complexes in vitro. Partial proteolysis reveals a smaller core consisting of the minimal death domain sequences. We have studied the tube/pelle interaction with the techniques of surface plasmon resonance, analytical ultracentrifugation and isothermal titration calorimetry. These measurements produce a value of K(d) for the complex of about 0.5 microM.

Probing activation of the prokaryotic arginine transcriptional regulator using chimeric proteins.

The major transcription factors controlling arginine metabolism in Escherichia coli and Bacillus subtilis, ArgR and AhrC, respectively, are homologous multimeric proteins that form l -arginine-dependent DNA-binding complexes capable of repressing transcription of the biosynthetic genes (both), activating transcription of catabolic genes (AhrC only) or facilitating plasmid dimer resolution (both). Multimerisation and l -arginine binding are associated with the C-terminal 70-80 residues; the N-terminal regions contain a winged helix-turn-helix DNA-binding domain. We have constructed chimeric genes in which the sequences for the N and C-terminal domains have been swapped. The resultant chimeric proteins and their corresponding native proteins have been analysed for their ability to multimerise and bind DNA operator sites in an L-arginine-dependent fashion. Gel filtration and equilibrium sedimentation analysis are consistent with the formation of hexamers by all four proteins in the presence of L-arginine and at high protein concentrations (>100 nM monomer). The hexamer sedimentation coefficients suggest that there is a reduction in molecular volume upon binding L-arginine, consistent with a conformational change accompanying an allosteric activation of DNA-binding. In the absence of L-arginine or at lower protein concentrations, the hexamers are clearly in rapid equilibrium with smaller subunits, whose dominant species appear to be based on trimers, as expected from the crystal structure of the ArgR C-terminal fragment, with the exception of the ArgR-C chimera, which apparently dissociates into dimers, suggesting that in the intact protein the DNA-binding domains may have a significant dimeric interaction. The hexamer-trimer Kdis in the micromolar range, suggesting that trimers are the principal species at in vivo concentrations.DNA binding by all four proteins has been probed by gel retardation and DNase I footprinting analysis using all three types of naturally occurring operators: biosynthetic sites encompassing two 18 bp ARG boxes separated by 2 bp; biosynthetic sites containing two such boxes and a third 18 bp ARG box at a distance of 100 bp downstream, i.e. within the structural gene; and finally a catabolic operator which contains a single ARG box site. The data show that all four proteins bind to the operators at the expected regions in an L-arginine-dependent fashion. From the apparent affinities of the chimeras for each target site, there is no obvious sequence-specificity associated with the N-terminal domains; rather the data can be interpreted in terms of differential allosteric activation, including DNA binding in the absence of L-arginine.Remarkably, the proteins show apparent "anti-competition" in the presence of excess, specific DNA fragments in gel retardation. This appears to be due to assembly of an activated form of the protein, probably hexamers, on the operator DNA. The data are discussed in terms of the current models for the mode of action of both native proteins.

Evidence of noncovalent dimerization of calmodulin.

Calcium-binding proteins, such as S-100, dimerize readily, and this phenomenon plays an important role in their regulation of target enzymes [Krebs, J., Quadroni, M. & Van Eldik, L.J. (1995) Nat. Struct. Biol. 2, 711-714; Kilby, P.M., Van Eldik, L.J. & Roberts, G. C. (1996) Structure 4, 1041-1052]. We have investigated by Fourier-transform ion cyclotron resonance (FTICR) MS the conformational states of the calcium-binding protein calmodulin, and present clear evidence for a calmodulin dimer formed as a result of noncovalent interactions between folded monomers. Ultra-high-resolution electrospray ionization (ESI) mass spectra for calmodulin, obtained with a 9.4 T FTICR mass spectrometer, are presented. With the use of denaturing solutions (1 : 1 acetonitrile/water + 1% formic acid), relatively high charge states (20 < z < 10) of monomeric calmodulin ions were detected, whereas when calmodulin was electrosprayed from buffer, monomers ions with only 5-10 charges were detected. CD measurements for calmodulin in buffered solution revealed that its alpha-helical content was significantly higher than that for calmodulin in acetonitrile/water solutions, consistent with a proposition that changes in charge state distributions observed in the MS experiments reflect differing states of calmodulin folding. Under buffered conditions, noncovalently bound calmodulin dimers were observed by ESI FTICR MS. Analytical ultracentrifugation experiments carried out in the same solution conditions as those used in the MS experiments were consistent with the proposed calmodulin dimer-monomer equilibrium. The ultra-high mass resolution achieved with the 9.4 T FTICR mass spectrometer allowed unequivocal identification of the noncovalent, as opposed to covalent, character of the calmodulin dimer.

The solution molecular weight and shape of the bacterial exopolysaccharides amylovoran and stewartan.

Amylovoran, the acidic exopolysaccharide (EPS) of Erwinia amylovora, and stewartan, the capsular EPS of E. stewartii, were characterized by analytical ultracentrifugation and by size exclusion chromatography connected to dual detection of light scattering and mass. The average molecular weights of amylovoran and stewartan were determined as 1.0 x 10(6) and 1.7 x 10(6) Da, with polydispersity values (Mw/M(n)) of 1.5 and 1.4, respectively. Based on the sugar composition and their molecular weight, both exopolysaccharides consist of approximately 1000 repeating units per molecule, this suggests a similar mechanism for chain length determination during biosynthesis of EPS in both organisms.

A polydisperse linear random coil model for the quaternary structure of pig colonic mucin.

The distribution of molecular weights for polymeric colonic mucus glycoprotein or "mucin" isolated and solubilised in the presence of protease inhibitors from pig colons is shown to be considerably greater than its "subunit" (thiol reduction product) and papain digested forms using the technique of size-exclusion chromatography coupled to multi-angle laser light scattering, and confirmed by sedimentation equilibrium measurements. The conformation of this mucin is probed by examining the molecular weight-intrinsic viscosity relationship in terms of the Mark-Houwink-Kuhn-Sakurada analysis for its polymeric (or "whole"), reduced and papain-digested forms: an exponent "a" of (1.1 +/- 0.1) is obtained indicating a linear random coil conformation consistent with other mucins. Size-exclusion chromatography coupled to multi-angle laser light scattering is shown to provide a relatively simple complementary technique to sedimentation equilibrium for the molecular weight distribution analysis of polydisperse materials.

A trimeric, alpha-helical, coiled coil peptide: association stoichiometry and interaction strength by analytical ultracentrifugation.

Alpha-helical coiled coils are proving to be almost ideal systems for the modelling of peptide and protein self-association processes. Stable oligomeric systems, in which the stoichiometry is well defined, can be produced by the careful selection of the appropriate amino acid sequence, although the principles behind this are still not fully understood. Here we report on a 35 residue peptide, FZ, synthesized by the solid phase method, which was originally designed to form a dimer, but which, in fact, associates to the trimeric state. A detailed characterization of the associative properties of the peptide has been performed by circular dichroism spectroscopy and, in particular, by sedimentation equilibrium in the analytical ultracentrifuge. The presence of the trimeric state, which is stable even at low peptide concentrations, has been confirmed by various independent methods of analysis for molar mass. The effects of both temperature and of guanidinium chloride on the peptide have been investigated and both found to be peptide-concentration dependent. The unfolding induced by the denaturant cannot be adequately described by a simple, two state monomer-trimer equilibrium.

Characterization of pig colonic mucins.

Pig colonic mucins isolated from the adherent mucus gel in the presence of proteinase inhibitors were solubilized by homogenization and the component mucins fractionated by CsC1 density-gradient centrifugation. Polymeric and reduced pig colonic mucin were both largely excluded on Sepharose CL-2B, papain-digested colonic mucin was included. The M(r) values of polymeric, reduced and digested mucins were 5.5 x 10(6), 2.1 x 10(6) and 0.6 x 10(6) respectively. This suggests that pig colonic mucin is comprised of 2-3 subunits, each subunit containing 3-4 glycosylated regions. The intrinsic viscosities of polymeric, reduced and digested mucin were 240 ml.g-1, 100 ml.g-1 and 20 ml.g-1 respectively. Polymeric pig colonic mucin comprised 16% protein per mg of glycoprotein and was rich in serine, threonine and proline (43% of total amino acids). There were approx. 150 disulphide bridges and 53 free thiol groups per mucin polymer. A seventh of the protein content was lost on reduction. This protein was particularly rich in proline and the hydrophobic amino acids. Papain-digested pig colonic mucin contained 11% protein per mg of glycoprotein and was rich in serine, threonine, glutamate and aspartate. All types of amino acids with the exception of aspartate were lost on digestion. The amino acid analysis of the proteolytically digested regions of pig colonic mucin are markedly different to the tandem repeat regions of the human mucin genes shown to be expressed in the colon.

Rapid size distribution and purity analysis of gastric mucus glycoproteins by size exclusion chromatography/multi angle laser light scattering.

The weight average molar masses and molar mass distributions of two commercial pig gastric mucins and two fresh mucin preparations were determined by the technique of size exclusion chromatography coupled on-line to multi angle laser light scattering (SEC/MALLS). Both commercial samples exhibited much lower molar mass averages than the freshly prepared material and contained more impurities. On the other hand, a fresh mucin preparation examined after being stored frozen for 18 months revealed a slight increase in molar mass.

Solution studies of the SH2 domain from the fyn tyrosine kinase: secondary structure, backbone dynamics and protein association.

The SH2 domain from Fyn tyrosine kinase, corresponding to residues 155-270 of the human enzyme, was expressed as a GST-fusion protein in a pGEX-E. coli system. After thrombin cleavage and removal of GST, the protein was studied by heteronuclear NMR. Two different phosphotyrosyl-peptides were synthesized and added to the SH2 domain. One peptide corresponded to the regulatory C-terminal tail region of Fyn. Sequence-specific assignment of NMR spectra was achieved using a combination of 1H-15N-correlated 2D HSQC, 15N-edited 3D TOCSY-HMQC, and 15N-edited 3D NOESY-HMQC spectra. By analysis of the alpha-proton chemical shifts and NOE intensities, the positions of secondary structural elements were determined and found to correspond closely to that seen in the crystal structure of the, homologous, Src-SH2 domain. To investigate the internal dynamics of the protein backbone, T1 and T2 relaxation parameters were measured on the free protein, as well as on both peptide complexes. Analytical ultracentrifugation and dynamic light scattering were employed to measure the effect of concentration and peptide-binding on self-association. The results suggest that, at NMR-sample concentrations, the free protein is present in at least dimeric form. Phosphopeptide binding and lower concentration significantly, but not completely, shift the equilibrium towards monomers. The possible role of this protein association in the regulation of the Src-family tyrosine kinases is discussed.