Side-chain conformations in 4-alpha-helical bundles.

The distribution of the chi(1), chi(2) dihedral angles in a dataset consisting of 12 unrelated 4-alpha-helical bundle proteins was determined and qualitatively compared with that observed in globular proteins. The analysis suggests that the 4-alpha-helical bundle motif could occasionally impose steric constraints on side chains: (i) the side-chain conformations are limited to only a subset of the conformations observed in globular proteins and for some amino acids they are sterically more constrained than those in helical regions of globular proteins; (ii) aspartic acid and asparagine occasionally adopt rotamers that have not been previously reported for globular or helical proteins; (iii) some rotamers of tyrosine and isoleucine are predominantly or exclusively associated with hydrophobic core positions (a, d); (iv) mutations in the hydrophobic core occur preferentially between residue types which among other physicochemical properties also share a predominant rotamer.     

Cytokine conformations: predictive studies.

The amino acid sequences of human and murine haemopoietins have been analysed using algorithms predictive for secondary structure. The results for 19 of these proteins (human and murine interleukins 2, 3, 4, 5, 6, 7 and granulocyte, macrophage and granulocyte macrophage-colony stimulating factors as well as human erythropoietin) suggest that they each contain a 4-alpha-helical bundle, ca 25 A long, as a common conformational feature. The most important predictive indicator was considered to be the occurrence of quasi-repeating sequences of seven amino acids of the form (a-b-c-d-e-f-g), with apolar side chains (usually leucine) lying alternately three and four residues apart in the a and d positions. As with other proteins of known secondary structure this periodicity favours the formation of alpha-helical elements, each with an apolar external strip, which interdigitate closely with one another when tested appropriately. Molecular models based on these putative 4-alpha-helical bundles are presented--with special reference to human granulocyte macrophage-colony stimulating factor. The extent to which such models are consistent with experiments designed to delineate receptor binding sites is discussed.     

Relationships between sequence and structure for the four-alpha-helix bundle tertiary motif in proteins.

The sequences of four-alpha-helical bundle proteins are characterized by a pattern of hydrophilic and hydrophobic amino acids which is repeated every seven residues. At each position of the heptad repeat there are specific constraints on the amino acid properties which result from the topology of the tertiary motif. These constraints give rise to patterns of amino acid distribution which are distinct from those of other proteins. The distributions in each of the heptad positions have been determined by a statistical analysis of structural and sequence data derived from seven families of aligned protein sequences. The constitution of each position is dominated by a very small number of different amino acids, with the core positions consisting overwhelmingly of Leu and Ala. The positional preferences of the individual amino acids can be generally interpreted in terms of residue properties and topological constraints. The potential for four-alpha-helix bundle folding is reflected primarily in the pattern of residue occurrence in the heptad and not in the overall amino acid composition of the protein. Possible applications of this analysis in structure predictions, sequence alignments and in the rational design and engineering of four-alpha-helical bundle proteins are discussed.     

Identification and characterization of functionally important elements in the multidrug resistance protein 1 COOH-terminal region

The ATP binding cassette (ABC) transporter, multidrug resistance protein 1 (MRP1/ABCC1), transports a broad spectrum of conjugated and unconjugated compounds, including natural product chemotherapeutic agents. In this study, we have investigated the importance of the COOH-terminal region of MRP1 for transport activity and basolateral plasma membrane trafficking. The COOH-terminal regions of some ABCC proteins have been implicated in protein trafficking, but the function of this region of MRP1 has not been defined. In contrast to results obtained with other ABCC proteins, we found that the COOH-proximal 30 amino acids of MRP1 can be removed without affecting trafficking to basolateral membranes. However, the truncated protein is inactive. Furthermore, removal of as few as 4 COOH-terminal amino acids profoundly decreases transport activity. Although amino acid sequence conservation of the COOH-terminal regions of ABC proteins is low, secondary structure predictions indicate that they consist of a broadly conserved helix-sheet-sheet-helix-helix structure. Consistent with a conservation of secondary and tertiary structure, MRP1 hybrids containing the COOH-terminal regions of either the homologous MRP2 or the distantly related P-glycoprotein were fully active and trafficked normally. Using mutated proteins, we have identified structural elements containing five conserved hydrophobic amino acids that are required for activity. We show that these are important for binding and hydrolysis of ATP by nucleotide binding domain 2. Based on crystal structures of several ABC proteins, we suggest that the conserved amino acids may stabilize a helical bundle formed by the COOH-terminal three helices and may contribute to interactions between the COOH-terminal region and the protein's two nucleotide binding domains.     

Two-dimensional NMR approaches to the study of protein structure and function

Two-dimensional Fourier transform methods for homonuclear proton NMR spectroscopy have been introduced by Wüthrich and Ernst as a means of extending assignments in spectra of proteins. Multinuclear two-dimensional approaches also appear promising. We are applying current one- and two-dimensional NMR methods to protein family members that differ from one another by one or more amino acid substitutions. The overall goal is to elucidate relationships among the sequences, structures, and functions of these proteins: for example, to delineate primary structural requirements for changes in observable properties such as conformation, amino acid side chain dynamics, hydrogen exchange dynamics, pK'a values, and oxidation-reduction potentials. The ovomucoids from a variety of species of birds, which include a single set of 12 pairs of third-domain proteins (Mr = 6062 for turkey third domain, similar for others) that differ by single amino acid substitutions, provide a favorable system for the study of the structural and dynamic effects of single amino acid replacements. X-ray crystallographic structures of two ovomucoid third domains are available. Other series of proteins being studied by these methods include the photosynthetic electron transport proteins ferredoxin and plastocyanin.     

Ab initio prediction of the three-dimensional structure of a de novo designed protein: a double-blind case study

Ab initio structure prediction and de novo protein design are two problems at the forefront of research in the fields of structural biology and chemistry. The goal of ab initio structure prediction of proteins is to correctly characterize the 3D structure of a protein using only the amino acid sequence as input. De novo protein design involves the production of novel protein sequences that adopt a desired fold. In this work, the results of a double-blind study are presented in which a new ab initio method was successfully used to predict the 3D structure of a protein designed through an experimental approach using binary patterned combinatorial libraries of de novo sequences. The predicted structure, which was produced before the experimental structure was known and without consideration of the design goals, and the final NMR analysis both characterize this protein as a 4-helix bundle. The similarity of these structures is evidenced by both small RMSD values between the coordinates of the two structures and a detailed analysis of the helical packing.     

Primary structure of a structural protein from the cuticle of the migratory locust, Locusta migratoria.

The complete amino acid sequence of a structural protein isolated from pharate cuticle of the locust Locusta migratoria was determined. The protein has an unusual amino acid composition: 42% of the residues are alanine and only 14 of the 20 common amino acid residues are present. The primary structure consists of regions enriched in particular amino acid residues. The N-terminal region and a region close to the C-terminus are enriched in glycine. The rest of the protein is dominated by alanine, except for two short regions enriched in hydrophilic residues. Almost all the proline residues are situated in the alanine-rich regions in a conserved sequence 'A-A-P-A/V'. An internal duplication has taken place covering most of the protein except for the glycine-rich regions. Owing to the unusual features of the protein a combination of automated Edman degradations and plasma-desorption m.s. was used to determine the complete sequence. The protein does not show sequence homology to other proteins, but proteins divided into regions enriched in the same kind of amino acid residues have been isolated from other insect structures.     

The OmpR-family of proteins: insight into the tertiary structure and functions of two-component regulator proteins

The Escherichia coli DNA-binding protein OmpR is the best characterized of those regulator proteins making up "two-component system," the simplest known form of bacterial signal transduction systems. Previous inspections of the E. coli genome DNA sequences have revealed that there are 15 proteins whose amino acid sequences show extensive similarities to that of OmpR (the OmpR-family of proteins). The three-dimensional structures of several OmpR-family proteins have been determined. In this review, we investigated the structures and amino acid sequences of this family of proteins. The results reveal several notable conservative varieties in their tertiary structures and functions.     

Structural identity between the iron- and manganese-containing superoxide dismutases

We have recently reported the first complete amino acid sequence of an iron-containing superoxide dismutase. The iron enzyme is thought to be closely homologous to the manganese-containing superoxide dismutases. The availability of complete amino acid sequence information for four manganese superoxide dismutases and the crystal structures for two iron and two manganese superoxide dismutases prompted us to investigate the degree of homology between the two proteins at various levels. We report that it is not possible to clearly distinguish the two proteins on the basis of their secondary or tertiary structures. It would appear that a small number of single site substitutions are responsible for conferring distinguishing properties between the two proteins. Substitution of glycine 77 and glutamine 154 by a glutamine and an alanine respectively in Photobacterium leiognathi iron superoxide dismutase may distinguish the kinetic and other particular properties of this protein from the manganese protein (and other iron superoxide dismutases). Furthermore the primary structure of both the iron and manganese proteins does not appear to have any homology with any other known amino acid sequence.     

Manipulation of the napin primary structure alters its packaging and deposition in transgenic tobacco (Nicotiana tabacum L.) seeds

Napin is a 2S storage protein found in the seeds of oilseed rape (Brassica napus L.) and related species. Using protein structural prediction programs we have identified a region in the napin protein sequence which forms a `hydrophilic loop' composed of amino acid residues located at the protein surface. Targeting this region, we have constructed two napin chimeric genes containing the coding sequence for the peptide hormone leucine-enkephalin as a topological marker. One version has a single enkephalin sequence of 11 amino acids including linkers and the second contains a tandem repeat of this peptide comprising 22 amino acids, inserted into the napin large subunit. The inserted peptide sequences alter the balance of hydrophilic to hydrophobic amino acids and introduce flexibility into this region of the polypeptide chain. The chimeric genes have been expressed in tobacco plants under the control of the seed-specific napA gene promoter. Analyses indicate that the engineered napin proteins are expressed, transported, post-translationally modified and deposited inside the protein bodies of the transgenic seeds demonstrating that the altered napin proteins behave in a similar fashion to the authentic napin protein. Detailed immunolocalisation studies indicate that the insertion of the peptide sequences has a significant effect on the distribution of the napin proteins within the tobacco seed protein bodies.     

Unusual features of cereal seed protein structure and evolution

The alcohol-soluble (prolamin) storage proteins of barley, wheat and rye vary in their structures, but all have two features in common: the presence of distinct structural domains differing in amino acid compositions, and of repeats within one of these domains. Detailed comparisons of amino acid sequences show that all appear to have evolved from a single ancestral gene consisting of three short related regions (called A, B and C). Regions related to A, B and C are also present in the minor prolamins of maize and in three other groups of seed proteins: inhibitors of alpha-amylase and/or trypsin from cereals. 25 storage globulins from several dicotyledonous species and a 2S albumin from sunflower. It is suggested that these proteins together constitute a protein superfamily with limited sequence homology.     

Nuclear magnetic resonance solution structure of PisI, a group B immunity protein that provides protection against the type IIa bacteriocin piscicolin 126, PisA

Lactic acid bacteria produce and secrete bacteriocins. These bacteriocins are potent antimicrobial peptides that are active against other closely related bacteria. As a means of self-protection, producer organisms also express immunity proteins. Immunity proteins are generally located on the same genetic locus and are cotranscribed with the bacteriocin. Although some cross immunity between bacteriocins has been observed, immunity proteins are typically highly specific. Immunity proteins for the type IIa bacteriocins range from 81 to 115 amino acids in length and display substantial variation in their sequences. Nonetheless, such immunity proteins have been classified into three groupings (groups A, B, and C) according to sequence homology. The structures of a group C (ImB2) and two group A (EntA-im and PedB) immunity proteins have previously been reported. We herein report the nuclear magnetic resonance solution structure of the remaining class of the type IIa immunity proteins. PisI, a 98-amino acid protein, is a group B immunity protein conferring immunity against piscicolin 126 (PisA). Like ImB2, EntA-im, and PedB, PisI folds into a globular protein in aqueous solution and contains an antiparallel four-helix bundle. Compared to ImB2 and EntA-im, PisI has a substantially longer and more flexible N-terminus, but a shorter C-terminus. No direct interaction between the bacteriocin and immunity protein is observed by NMR in either aqueous or membrane mimicking environments. This further suggests that the mechanism that mediates immunity is not due to a direct bacteriocin-immunity protein interaction but rather is receptor-mediated. It has now been confirmed that the four-helix bundle is indeed a structural motif among the type IIa immunity proteins.     

Identification of a stability determinant on the edge of the Tet repressor four-helix bundle dimerization motif

Isofunctional tetracycline repressor (TetR) proteins isolated from different bacteria show a sequence identity between 38 and 88% of the residues. Their active state is a homodimer formed by a four-alpha-helix bundle as the main interaction motif. We utilize this sequence variation of isofunctional proteins to determine residues contributing to the stability of the four-helix bundle. The thermodynamic stabilities of two TetR proteins with 63% sequence identity were determined by urea-induced reversible denaturation followed by fluorescence and circular dichroism. Both methods yield identical results. The deltaG(o)U (H2O) values are 60 and 75 kJ x mol(-1). We have constructed TetR hybrid proteins derived from these wild types to identify the determinant leading to the 15 kJ x mol(-1) stability difference. Successive size reduction of the exchanged portion yielded two single residues affecting the overall protein stability. The P184Q exchange leads to a more stable protein, whereas the G181D exchange located at the solvent's exposed edge of the four-helix bundle is solely responsible for the reduced stability. Additional mutants based on crystal structures of TetR do not reveal any hint for steric interference of the Asp181 side chain with neighboring residues. Thus, this is an example for the role played by surface-exposed turn residues for the stability of four-helix bundles. We assume that the larger conformational flexibility of Gly and the reduction of the negative surface charge could favor formation of the turn on the edge of the four-helix bundle.     

A topological model for the haemolysin translocator protein HlyD

Summary A topological model for HlyD is proposed that is based on results obtained with gene fusions of lacZ and phoA to hlyD. Active H1yD-LacZ fusion proteins were only generated when lacZ was fused to hlyD. within the first 180 by (60 amino acids). H1yD-PhoA proteins exhibiting alkaline phosphatase (AP) activity were obtained when phoA was inserted into hlyD. between nucleotides 262 (behind amino acid position 87) and 1405 (behind amino acid position 468, only 10 amino acids away from the C-terminus of HlyD Active insertions of phoA into the middle region of hlyD. were not observed on in vivo transposition but such fusions exhibiting AP activity could be constructed by in vitro techniques. A fusion protein that carried the PhoA part close to the C-terminal end of HlyD proved to be the most stable HlyD-PhoA fusion protein. In contrast to the other, rather unstable, HlyD-PhoA⁺ fusions, no proteolytic degradation product of this HlyD-PhoA protein was observed and nearly all the alkaline phosphatase activity was membrane bound. Protease accessibility and cell fractionation experiments indicated that the alkaline phosphatase moiety of this fusion protein was located in the periplasm as for all other HlyD-PhoA⁺ proteins. These data and computer-assisted predictions suggest a topological model for HlyD with the N-terminal 60 amino acids located in the cytoplasm, a single transmembrane segment from amino acids 60 to 80 and a large periplasmic region extending from amino acid 80 to the C-terminus. Neither the HlyD fusion proteins obtained nor a mutant HlyD protein that had lost the last 10 amino acids from the C-terminus of HlyD exhibited translocator activity for HlyA or other reporter proteins carrying the HlyA signal sequence. The C-terminal 10 amino acids of HlyD showed significant similarity with the corresponding sequences of other HlyD-related proteins involved in protein secretion.     

Probing protein fold space with a simplified model

We probe the stability and near-native energy landscape of protein fold space using powerful conformational sampling methods together with simple reduced models and statistical potentials. Fold space is represented by a set of 280 protein domains spanning all topological classes and having a wide range of lengths (33-300 residues) amino acid composition and number of secondary structural elements. The degrees of freedom are taken as the loop torsion angles. This choice preserves the native secondary structure but allows the tertiary structure to change. The proteins are represented by three-point per residue, three-dimensional models with statistical potentials derived from a knowledge-based study of known protein structures. When this space is sampled by a combination of parallel tempering and equi-energy Monte Carlo, we find that the three-point model captures the known stability of protein native structures with stable energy basins that are near-native (all α: 4.77 Å, all β: 2.93 Å, α/β: 3.09 Å, α+β: 4.89 Å on average and within 6 Å for 71.41%, 92.85%, 94.29% and 64.28% for all-α, all-β, α/β and α+β, classes, respectively). Denatured structures also occur and these have interesting structural properties that shed light on the different landscape characteristics of α and β folds. We find that α/β proteins with alternating α and β segments (such as the β-barrel) are more stable than proteins in other fold classes.     

Topological comparison of two helix destabilizing proteins: ribonuclease A and the gene 5 DNA binding protein

A topological comparison of the two helix destabilizing proteins, pancreatic ribonuclease A and the gene 5 DNA binding protein of bacteriophage fd has been completed utilizing the available high resolution tertiary structures of each protein. The results indicate these two proteins are structurally if not also evolutionarily related. Regions of closet topological equivalence occur between beta loops directly involved in nucleotide binding or are required for the maintenance of their respective oligonucleotide binding channels. In addition, there is a similar placement of critical amino acid side chains about the binding site. Further evidence for this structural relationship is obtained by comparison of structural data for the mode of complexation of polynucleotides to each protein. The results of topological comparison suggest the essential property shared by helix destabilizing proteins, whether specialized DNA binding proteins such as G5BP or proteins with other primary functional roles, like ribonuclease A, is the presence of an elongated oligonucleotide binding channel. Although ribonuclease A and G5BP are structurally related, it seems likely any protein with this structural feature will exhibit a helix destabilizing capacity. This conclusion is supported by the diversity of molecular characteristics shown by other proteins having this activity.     

Six-helix bundle assembly and analysis of the central core of mumps virus fusion protein

The fusion protein of enveloped viruses mediates the fusion between the viral and cellular membranes, allowing the penetration of the viral genomes into the host cell. Many of these proteins share a common fold comprising a central core trimer of anti-parallel coiled-coil heterodimers, which are formed by two discontinuous heptad repeat (HR) motifs located at the ectodomain of the fusion proteins. In this study, we constructed and purified the corresponding regions (HR1 and HR2) of mumps virus fusion protein that are predicted to form coiled coil. The HR1 and HR2 were expressed and purified separately or as a single chain connected by an amino acid linker (HR1-linker-HR2, named 2-Helix). Series of biochemical and biophysical analyses of the expressed proteins have shown that HR1 and HR2 of mumps virus fusion protein share the common features of other enveloped virus fusion proteins. CD spectral results show that HR1 forms an alpha-helical coil structure while HR2 exists as an unstructured monomer in PBS in nature. Mixtures of HR1 and HR2 could form a stable six-helix bundle, indicating the interaction of HR1 and HR2. The 2-Helix protein also shows characteristic properties of the 6-helix bundle. Therefore, mumps virus fusion protein has a common core architecture and its HR regions could be used as a drug target for virus fusion inhibitors.     

Identification of four major hornet silk genes with a complex of alanine-rich and serine-rich sequences in Vespa simillima xanthoptera Cameron

Hornet silk, a fibrous protein in the cocoon produced by the larva of the vespa, is composed of four major proteins. In this study, we constructed silk-gland cDNA libraries from larvae of the hornet Vespa simillima xanthoptera Cameron and deduced the full amino acid sequences of the four hornet silk proteins, which were named Vssilk 1-4 in increasing order of molecular size. Portions of the amino acid sequences of the four proteins were confirmed by Matrix-assisted laser desorption/ionization-time of flight/mass spectrometry (MALDI-TOF/MS) and N-terminal protein sequencing. The primary sequences of the four Vssilk proteins (1-4) were highly divergent, but the four proteins had some common properties: (i) the amino acid compositions of all four proteins were similar to each other in that the well-defined and characteristic repetitive patterns present in most of the known silk proteins were absent; and (ii) the characteristics of the amino acid sequences of the four proteins were also similar in that Ser-rich structures such as sericin were localized at both ends of the chains and Ala-rich structures such as fibroin were found in the center. These characteristic primary structures might be responsible for the coexisting alpha-helix and beta-sheet conformations that make up the unique secondary structure of hornet silk proteins in the native state. Because heptad repeat sequences of hydrophobic residue are present in the Ala-rich region, we believe that the Ala-rich region of hornet silk predominantly forms a coiled coil with an alpha-helix conformation.     

Evaluation of the protein interfaces that form an intermolecular four-helix bundle as studied by computer simulation

Rational design of protein surface is important for creating higher order protein structures, but it is still challenging. In this study, we designed in silico the several binding interfaces on protein surfaces that allow a de novo protein–protein interaction to be formed. We used a computer simulation technique to find appropriate amino acid arrangements for the binding interface. The protein–protein interaction can be made by forming an intermolecular four-helix bundle structure, which is often found in naturally occurring protein subunit interfaces. As a model protein, we used a helical protein, YciF. Molecular dynamics simulation showed that a new protein–protein interaction is formed depending on the number of hydrophobic and charged amino acid residues present in the binding surfaces. However, too many hydrophobic amino acid residues present in the interface negatively affected on the binding. Finally, we found an appropriate arrangement of hydrophobic and charged amino acid residues that induces a protein–protein interaction through an intermolecular four-helix bundle formation.