Crystal structures of Scone: pseudosymmetric folding of a symmetric designer protein.

Recent years have seen an increase in the development of computational proteins, including symmetric ones. A ninefold-symmetric ß-propeller protein named Cake has recently been developed. Here, attempts were made to further engineer this protein into a threefold-symmetric nine-bladed propeller using computational design. Two nine-bladed propeller proteins were designed, named Scone-E and Scone-R. Crystallography, however, revealed the structure of both designs to adopt an eightfold conformation with distorted termini, leading to a pseudo-symmetric protein. One of the proteins could only be crystallized upon the addition of a polyoxometalate, highlighting the usefulness of these molecules as crystallization additives.

The structural basis of sequence-independent peptide binding by OppA protein.

Specific protein-ligand interactions are critical for cellular function, and most proteins select their partners with sharp discrimination. However, the oligopeptide-binding protein of Salmonella typhimurium (OppA) binds peptides of two to five amino acid residues without regard to sequence. The crystal structure of OppA reveals a three-domain organization, unlike other periplasmic binding proteins. In OppA-peptide complexes, the ligands are completely enclosed in the protein interior, a mode of binding that normally imposes tight specificity. The protein fulfills the hydrogen bonding and electrostatic potential of the ligand main chain and accommodates the peptide side chains in voluminous hydrated cavities.

What is the true structure of liganded haemoglobin?

Does the crystal structure of a protein accurately represent its structure in solution? Or does the crystallization process perturb the structure significantly? Although aware of the problem, most crystallographers would argue that the highly solvated and weakly held lattice in protein crystals is, in general, unlikely to shift ordered parts of the molecule. In the case of conformationally flexible proteins, however, there is the possibility that one form might be favoured over another. Several lines of evidence suggest that this might be the case for the crystal structure of liganded Hb, although conflicting data exist.

The crystal structures of the oligopeptide-binding protein OppA complexed with tripeptide and tetrapeptide ligands.

BACKGROUND: The periplasmic oligopeptide-binding protein OppA has a remarkably broad substrate specificity, binding peptides of two or five amino-acid residues with high affinity, but little regard to sequence. It is therefore an ideal system for studying how different chemical groups can be accommodated in a protein interior. The ability of the protein to bind peptides of different lengths has been studied by co-crystallising it with different ligands. RESULTS: Crystals of OppA from Salmonella typhimurium complexed with the peptides Lys-Lys-Lys (KKK) and Lys-Lys-Lys-Ala (KKKA) have been grown in the presence of uranyl ions which form important crystal contacts. These structures have been refined to 1.4 A and 2.1 A, respectively. The ligands are completely enclosed, their side chains pointing into large hydrated cavities and making few strong interactions with the protein. CONCLUSIONS: Tight peptide binding by OppA arises from strong hydrogen bonding and electrostatic interactions between the protein and the main chain of the ligand. Different basic side chains on the protein form salt bridges with the C terminus of peptide ligands of different lengths.

Proteolytic analysis of the FliH/FliI complex, the ATPase component of the type III flagellar export apparatus of Salmonella.

The ATPase FliI of the Salmonella type III flagellar protein export apparatus is a 456 amino acid residue cytoplasmic protein consisting of two regions, an N-terminal flagellum-specific region and a C-terminal ATPase region. It forms a complex with a regulatory protein FliH in the cytoplasm. Multi-angle light-scattering studies indicate that FliH forms a homodimer, (FliH)2, and that FliH and FliI together form a heterotrimer, (FliH)2FliI. Mobility upon gel-filtration chromatography gives much higher apparent molecular masses for both species, whereas the mobility of FliI is normal. Sedimentation velocity measurements indicate that both (FliH)2 and the FliH/FliI complex are quite elongated. We have analyzed FliH, FliI and the FliH/FliI complex for proteolytic sensitivity. FliI was degraded by clostripain into two stable fragments, one of 48 kDa (FliI(CL48), missing the first seven amino acid residues) and the other of 46 kDa (FliI(CL46), missing the first 26 residues). Small amounts of two closely spaced 38 kDa fragments (FliI(CL38), missing the first 93 and 97 residues, respectively) were also detected. The FliH homodimer was insensitive to clostripain proteolysis and provided protection to FliI within the FliH/FliI complex. Neither FliI(CL48) nor FliI(CL46) could form a complex with FliH, demonstrating that the N terminus of FliI is essential for the interaction. ATP, AMP-PNP, and ADP bound forms of FliI within the FliH/FliI complex regained sensitivity to clostripain cleavage. Also, the sensitivity of the two FliI(CL38) cleavage sites was much greater in the ATP and AMP-PNP bound forms than in either the ADP bound form or nucleotide-free FliI. The ATPase domain itself was insensitive to clostripain cleavage. We suggest that the N-terminal flagellum-specific region of FliI is flexible and changes its conformation during the ATP hydrolysis cycle.

The crystal structures of trout Hb I in the deoxy and carbonmonoxy forms.

We have determined the X-ray crystallographic structure of trout Hb I in both the deoxy and carbonmonoxy forms to resolution limits of 2.3 angstroms and 2.5 angstroms, respectively. The overall fold of the molecule is highly similar to that of human HbA despite the low level of sequence identity between these proteins. Trout Hb I is unusual in displaying almost no pH dependence of oxygen binding affinity, and (at most) very weak interactions with heterotropic effector ligands such as organic phosphates. Comparison of the two quaternary states of the protein indicates how such effects are minimised and how the low-affinity T state of the protein is stabilised in the absence of heterotropic interactions.

The crystal structure of a high oxygen affinity species of haemoglobin (bar-headed goose haemoglobin in the oxy form).

We have determined the crystal structure of bar-headed goose haemoglobin in the oxy form to a resolution of 2.0 A. The R-factor of the model is 19.8%. The structure is similar to human HbA, but contacts between the subunits show slightly altered packing of the tetramer. Bar-headed goose blood shows a greatly elevated oxygen affinity compared to closely related species of geese. This is apparently due to a single proline to alanine mutation at the alpha 1 beta 1 interface which destabilises the T state of the protein. The beta chain N and C termini are well-localized, and together with other neighbouring basic groups they form a strongly positively charged groove at the entrance to the central cavity around the molecular dyad. The well-ordered conformation and the three-dimensional distribution of positive charges clearly indicate this area to be the inositol pentaphosphate binding site of bird haemoglobins.

Crystallographic and calorimetric analysis of peptide binding to OppA protein.

Isothermal titration calorimetry has been used to study the binding of 20 different peptides to the peptide binding protein OppA, and the crystal structures of the ligand complexes have been refined. This periplasmic binding protein, part of the oligopeptide permease system of Gram negative bacteria, has evolved to bind and enclose small peptides of widely varying sequences. The peptides used in this study have the sequence Lys-X-Lys, where X is any of the 20 commonly occurring amino acids. The various side-chains found at position 2 on the ligand fit into a hydrated pocket. The majority of side-chains are restrained to particular conformations within the pocket. Water molecules act as flexible adapters, matching the hydrogen-bonding requirements of the protein and ligand and shielding charges on the buried ligand. This use of water by OppA to broaden the repertoire of its binding site is not unique, but contrasts sharply with other proteins which use water to help bind ligands highly selectively. Predicting the thermodynamics of binding from the structure of the complexes is highly complicated by the influence of water on the system.

Specificity and interactions of the protein OppA: partitioning solvent binding effects using mass spectrometry.

Mass spectrometry (MS) was used to characterise the binding of the 58 kDa protein OppA to 11 peptides with diverse properties. Peptides with two, three and five amino acid residues were added to OppA, and the mass spectra showed that the highest-affinity complexes are formed between OppA and tripeptide ligands. Lower-affinity complexes were observed for OppA and dipeptide ligands, and no complex formation was detected with pentapeptides or a tripeptide in which the N-terminal amino group was acetylated. Tripeptides containing a single d amino acid residue were found not to bind to native OppA. Evidence from the peak width and the, charge in the spectra of the complexes suggests that the bound peptides are encapsulated by the protein in a solvent-filled cavity in the gas phase of the mass spectrometer. Analysis of the proportions of peptide-bound and free proteins under low-energy MS conditions shows a good correlation with solution-phase K(d) measurements where available. Increasing the internal energy of the gas-phase complex led to dissociation of the complex. The ease of dissociation is interpreted in terms of the intrinsic stability of the complex in the absence of the stabilising effects of bulk solvent. The results from this study demonstrate insensitivity to the hydrophobic and ionic properties, of the side-chains of the peptides, in contrast to the investigation of other protein ligand systems by MS. Moreover, these findings are in accord with the physiological role of this protein in allowing into the cell di- and tripeptides containing naturally occurring amino acids, regardless of their sequence, while barring access to potentially harmful peptide mimics.

The functional similarity and structural diversity of human and cartilaginous fish hemoglobins.

Although many descriptions of adaptive molecular evolution of vertebrate hemoglobins (Hb) can be found in physiological text books, they are based mainly on changes of the primary structure and place more emphasis on conservation than alterations at the functional site. Sequence analysis alone, however, does not reveal much about the evolution of new functions in proteins. It was found recently that there are many functionally important structural differences between human and a ray (Dasyatis akajei) Hb even where sequence is conserved between the two. We have solved the structures of the deoxy and CO forms of a second cartilaginous fish (a shark, Mustelus griseus) Hb, and compared it with structures of human Hb, two bony fish Hbs and the ray Hb in order to understand more about how vertebrate Hbs have functionally evolved by the selection of random amino acid substitutions. The sequence identity of cartilaginous fish Hb and human Hb is a little less than 40 %, with many functionally important amino acid replacements. Wider substitutions than usually considered as neutral have been accepted in the course of molecular evolution of Hb. As with the ray Hb, the shark Hb shows functionally important structural differences from human Hb that involve amino acid substitutions and shifts of preserved amino acid residues induced by substitutions in other parts of the molecule. Most importantly, beta E11Val in deoxy human Hb, which overlaps the ligand binding site and is considered to play a key role in controlling the oxygen affinity, moves away about 1 A in both the shark and ray Hbs. Thus adaptive molecular evolution is feasible as a result of both functionally significant mutations and deviations of preserved amino acid residues induced by other amino acid substitutions.

Relating structure to thermodynamics: the crystal structures and binding affinity of eight OppA-peptide complexes.

The oligopeptide-binding protein OppA provides a useful model system for studying the physical chemistry underlying noncovalent interactions since it binds a variety of readily synthesized ligands. We have studied the binding of eight closely related tripeptides of the type Lysine-X-Lysine, where X is an abnormal amino acid, by isothermal titration calorimetry (ITC) and X-ray crystallography. The tripeptides fall into three series of ligands, which have been designed to examine the effects of small changes to the central side chain. Three ligands have a primary amine as the second side chain, two have a straight alkane chain, and three have ring systems. The results have revealed a definite preference for the binding of hydrophobic residues over the positively charged side chains, the latter binding only weakly due to unfavorable enthalpic effects. Within the series of positively charged groups, a point of lowest affinity has been identified and this is proposed to arise from unfavorable electrostatic interactions in the pocket, including the disruption of a key salt bridge. Marked entropy-enthalpy compensation is found across the series, and some of the difficulties in designing tightly binding ligands have been highlighted.

Purification of the autotransporter protein Hbp of Escherichia coli.

The enzyme Hbp (hemoglobin protease) of the pathogenic Escherichia coli strain EB1 has been purified to homogeneity by gel filtration chromatography. The purified protein is capable of binding heme and shows hemoglobin protease activity. Our method of purification is applicable not only to Hbp but also to other autotransporter proteins and will contribute to a better understanding of the function-structure relationship of this family of proteins.

Scoring functions: a view from the bench.

Computational approaches to drug design are presently hindered by the complexity of the physical chemistry which underlies weak, non-covalent interactions between protein targets and small molecule ligands. Although a number of programs are now available for the design of novel potential ligands, it remains a key problem to rank these rapidly and reliably by estimated binding affinity. Such a step is necessary to select only the most promising candidates for synthesis and experimental characterisation. To calculate ligand affinity quickly and reliably is an extremely difficult problem, but it may well prove possible to estimate sufficiently accurately given an appropriate set of parameters to 'score' individual protein-ligand interactions. Improvements in the situation will require a wider set of thermodynamically characterised systems than is currently available.

Ab initio phasing of a 4189-atom protein structure at 1.2 A resolution.

The phase problem remains a key rate-limiting step in the determination of macromolecular X-ray structures. Direct methods, applying probability theory to the native data set, can routinely solve structures of up to about 200 non-H atoms, although much larger structures have been solved given sufficiently high resolution data and the presence of heavy atoms. Here it is shown that maximum-likelihood refinement of free-atom models with ARP/wARP can solve ab initio a much larger metalloprotein structure than the largest so far solved by conventional direct methods. The protein, OppA, is not naturally associated with metal ions but was co-crystallized with uranium.

The structures of deoxy human haemoglobin and the mutant Hb Tyralpha42His at 120 K.

The structures of deoxy human haemoglobin and an artificial mutant (Tyralpha42-->His) have been solved at 120 K. While overall agreement between these structures and others in the PDB is very good, certain side chains are found to be shifted, absent from the electron-density map or in different rotamers. Non-crystallographic symmetry (NCS) is very well obeyed in the native protein, but not around the site of the changed residue in the mutant. NCS is also not obeyed by the water molecule invariably found in the alpha-chain haem pocket in room-temperature crystal structures of haemoglobin. At 120 K, this water molecule disappears from one alpha chain in the asymmetric unit but not the other.

Peptide binding in OppA, the crystal structures of the periplasmic oligopeptide binding protein in the unliganded form and in complex with lysyllysine.

The periplasmic oligopeptide binding protein, OppA, acts as the initial receptor for the uptake of peptides by the oligopeptide permease (Opp) in Gram-negative bacteria. Opp will handle peptides between two and five amino acid residues regardless of their sequence. The crystal structures of a series of OppA-peptide complexes have revealed an enclosed but versatile peptide binding pocket and have illustrated how tri- and tetrapeptide ligands are accommodated. Here, the crystal structures of (i) OppA complexed with a dipeptide (lysyllysine) and (ii) unliganded OppA have been solved using X-ray data extending to 1.8 and 2.4 A spacing, respectively. In the dipeptide complex, the alpha-amino group of the ligand is anchored through an ion pair interaction with Asp419, as observed in complexes with longer peptides. However, its alpha-carboxylate group forms water-mediated interactions with the guanidinium groups of Arg404 and Arg413 rather than the direct salt bridges to Arg413 and His371 observed in the tripeptide and tetrapeptide complexes, respectively. Isothermal titration calorimetric measurements of the binding of lysine-containing peptides of different lengths to OppA show that the dipeptide, KK, is bound with approximately 60-fold lower affinity than related tri- and tetrapeptides (KKK and KKKA, respectively). These data are discussed with reference to the calculated enthalpic and entropic contributions to ligand binding and the structures of the OppA peptide complexes. In the unliganded molecule, domain III has rotated as a rigid body through 26 degrees away from domains I and II, exposing the ligand binding site. The water structure in the binding cleft shows similarities to that in the various OppA-peptide complexes.

Structure determination of OppA at 2.3 A resolution using multiple-wavelength anomalous dispersion methods.

OppA is a 58.8 kDa bacterial transport protein involved in the transport of peptides across the cytoplasmic membrane of Gram-negative bacteria. It binds peptides from two to five residues in length but with little sequence specificity. OppA from Salmonella typhimurium has been cloned and expressed in E. coli and the protein cocrystallized with uranyl acetate, producing two distinct crystal forms with different uranium sites. Multiple-wavelength data collected about the uranium L(III) edge have been collected at the Daresbury Synchrotron Radiation Source (SRS) to a nominal resolution limit of 2.3 A. Maximum-likelihood phasing methods have been used in phase determination from the multiple-wavelength data giving a readily interpretable electron-density map, without any density modification. The electron-density map, calculated at 2.3 A resolution shows OppA to be a bilobal, principally beta-stranded, three-domain protein. The tri-lysine ligand molecule can be clearly seen in the peptide-binding site between the two lobes.

The dimerization interface of the metastasis-associated protein S100A4 (Mts1): in vivo and in vitro studies.

The S100 calcium-binding proteins are implicated in signal transduction, motility, and cytoskeletal dynamics. The three-dimensional structure of several S100 proteins revealed that the proteins form non-covalent dimers. However, the mechanism of the S100 dimerization is still obscure. In this study we characterized the dimerization of S100A4 (also named Mts1) in vitro and in vivo. Analytical ultracentrifugation revealed that apoS100A4 was present in solution as a mixture of monomers and dimers in a rapidly reversible equilibrium (K(d) = 4 +/- 2 microm). The binding of calcium promoted dimerization. Replacement of Tyr-75 by Phe resulted in the stabilization of the dimer. Helix IV is known to form the major part of the dimerization interface in homologous S100 proteins. By using the yeast two-hybrid system we showed that only a few residues of helix IV, namely Phe-72, Tyr-75, Phe-78, and Leu-79, are essential for dimerization in vivo. A homology model demonstrated that these residues form a hydrophobic cluster on helix IV. Their role is to stabilize the structure of individual subunits rather than provide specific interactions across the dimerization surface. Our mutation data showed that the specificity at the dimerization surface is not particularly stringent, which is consistent with recent data indicating that S100 proteins can form heterodimers.