Alignment of protein sequences using the hydrophobic core scores

Making an alignment of the amino acid sequences is an essential step in the prediction of an unknown protein structure by model building from the known structure of a protein of the same family. To improve the accuracy of the alignments, we introduced the concept of hydrophobic core scores, which restrains putting insertions/deletions in the hydrophobic core regions of the protein. Eight pairs of protein sequences were aligned by this method, and the quality of the alignments were assessed by reference to those obtained by the structural superposition. The introduction of the hydrophobic core scores derived from the knowledge of the tertiary structure of one of each pair resulted in an improvement of the accuracy of the alignments. The quality of the alignment was found to depend on the homology of the protein sequences.     

Mechanisms controlling steroid receptor binding to specific DNA sequences.

Mammalian progesterone receptors activated by hormone binding in nuclei of intact cells exhibit substantially higher binding activity for specific DNA sequences than receptors bound with hormone and activated in cell-free cytosol. Differences in DNA-binding activity occur despite the fact that both activated receptor forms sediment at 4S on sucrose gradients and are apparently dissociated from the heat shock protein 90. This suggests that hormone-induced release of heat shock protein 90 from receptors is necessary, but not sufficient for maximal activation of DNA binding. This report is a review of studies from our laboratories that have examined the role of receptor interaction with other nuclear protein factor(s), and receptor dimerization in solution, as additional regulatory steps involved in the process of receptor activation and binding to specific gene sequences.     

Conformation-function relationships in hydrophobic peptides: interior turns and signal sequences

Two studies are diescribed in which synthetic peptides have been designed and examined to address biochemical problems inherent in hydorphobic environments: (1) The cyclic hexapeptide cyclo-(D -Tyr(Bzl)-Gly-Ile-Leu-Gln-Pro) was synthesized as a model of an interior β-turn from the protein lysozyme. Conformational analysis by proton nmr methods, including two-dimensional nulcear Overhauser effect spectroscopy, revealed that the model peptide adopts one conformation in chloroform/dimethyl sulfoxide (98.2) and tetramethylene sulfone solutions. The conformation consists of two linked β-turns, one with the same sequence (Gly-Ile-Leu-Gln) and geometry (Type I) as the protein turn. (2) Major portions of the λ-receptor protein (LamB) signal sequences from E. coli wildtype and mutant strains have been synthesized. The conformational properties and membrane interactions of these synthetic signal peptides correlate with the in vivo export function of the wild type and mutant strains. Functional signal sequences are significantly richer in α-helix in aaqueous trifluoroethanol, lysolecithin, or sodium do-decyl sulfate solution than is a nonfunctional mutant signal sequence.     

Nuclear protein factors binding with specific DNA sequences

Primary structure of thousands of genes is being determined in many laboratories worldwide. While it is relatively easy to analyse the coding region(s) of genes, it is usually hard to understand what is located in non-coding regions. A non-coding region may contain very valuable information about the mode of functioning of a given gene, e. g. promoters, enhancers, silencers etc. The regulatory function of these sequences is determined by their interaction with certain sequence-specific proteins, i. e. the presence of a certain DNA sequence in a non-coding region of a gene may suggest that the gene is regulated by a specific protein factor. This minireview summarizes recent data on most known eukaryotic sequence-specific DNA-binding protein factors, including their origin, DNA consensus, and their role in expression of corresponding genes.     

Preferential binding to DNA sequences of peptides related to a novel XPRK motif

Two dodecapeptide amines: (WPRK)(3)NH(2)[WR-12] and (YPRK)(3)NH(2)[YR-12], and a 30-mer polypeptide amide (SP-30) were synthesized by solid-phase peptide methodology. DNase I footprinting studies on a 117-mer DNA showed that WR-12 and YR-12 bind selectively to DNA sequences in a manner similar to SP-30 which has a repeating SPK(R)K sequence. The most distinctive blockages seen with all three peptides occur at positions 26-30, 21-24 and 38-45 around sequences 5'-GAATT-3', 5'-TAAT-3' and 5'-AAAACGAC-3', respectively. However, it appears that YR-12 is better able to extend its recognition site to include CG pairs than is SP-30. At low concentrations YR-12 was able to induce enhanced rates of DNase I cleavage at regions surrounding some of its binding sites. To obtain further quantitative data supplementary to the footprinting work, equilibrium binding experiments were performed in which the binding of the two peptides to six decanucleotide duplexes was compared. Scatchard analyses indicated that WR-12 may be more selective for oligomers containing runs of consecutive purines or pyrimidines. On the other hand, YR-12 binds better to d(CTTAGACGTC)- d(GACGTCTAAG) than to the other oligomer duplexes, denoting selectivity for evenly distributed C/G and A/T sequences.     

The presence of a helix breaker in the hydrophobic core of signal sequences of secretory proteins prevents recognition by the signal-recognition particle in Escherichia coli.

Signal sequences often contain alpha-helix-destabilizing amino acids within the hydrophobic core. In the precursor of the Escherichia coli outer-membrane protein PhoE, the glycine residue at position -10 (Gly-10) is thought to be responsible for the break in the alpha-helix. Previously, we showed that substitution of Gly-10 by alpha-helix-promoting residues (Ala, Cys or Leu) reduced the proton-motive force dependency of the translocation of the precursor, but the actual role of the helix breaker remained obscure. Here, we considered the possibility that extension of the alpha-helical structure in the signal sequence resulting from the Gly-10 substitutions affects the targeting pathway of the precursor. Indeed, the mutations resulted in reduced dependency on SecB for targeting in vivo. In vitro cross-linking experiments revealed that the G-10L and G-10C mutant PhoE precursors had a dramatically increased affinity for P48, one of the constituents of the signal-recognition particle (SRP). Furthermore, in vitro cross-linking experiments revealed that the G-10L mutant protein is routed to the SecYEG translocon via the SRP pathway, the targeting pathway that is exploited by integral inner-membrane proteins. Together, these data indicate that the helix breaker in cleavable signal sequences prevents recognition by SRP and is thereby, together with the hydrophobicity of the signal sequence, a determinant of the targeting pathway.     

The entire N-terminal half of TatC is involved in twin-arginine precursor binding

Translocation of twin-arginine precursor proteins across the cytoplasmic membrane of Escherichia coli requires the three membrane proteins TatA, TatB, and TatC. TatC and TatB were shown to be involved in precursor binding. We have analyzed in vitro a number of single alanine substitutions in tatC that were previously shown to compromise in vivo the function of the Tat translocase. All tatC mutants that were defective in precursor translocation into cytoplasmic membrane vesicles concomitantly interfered with precursor binding not only to TatC but also to TatB. Hence structural changes of TatC that affect precursor targeting simultaneously abolish engagement of the twin-arginine signal sequence with TatB and block the formation of a functional Tat translocase. Since these phenotypes were observed for tatC mutations spread over the first half of TatC, this entire part of the molecule must globally be involved in precursor binding.     

Parallel effects of signal peptide hydrophobic core modifications on co-translational translocation and post-translational cleavage by purified signal peptidase

The length of the hydrophobic core of the bovine parathyroid hormone signal peptide was modified by in vitro mutagenesis. Extension of the hydrophobic core by three amino acids at the NH2-terminal end had little effect on the proteolytic processing of the signal peptide by microsomal membranes. Deletion of 6 of the 12 amino acids in the core eliminated translocation and processing of the modified protein. Deletion of pairs of amino acids across the core resulted in position-dependent inhibition of signal activity unrelated to hydrophobicity but inversely related to the hydrophobic moments of the modified cores. Deletions in the NH2-terminal region of the core were strongly inhibitory for proteolytic processing whereas deletions in the COOH-terminal region had no effect or increased processing when assessed either co-translationally with microsomal membranes or post-translationally with purified hen oviduct signal peptidase. Deletion of cysteine 18 and alanine 19 increased processing, but deletion of cysteine alone or substitution of leucine for cysteine did not increase processing more than deletion of both residues at 18 and 19. Translations of the translocation-defective mutants with pairs of amino acids deleted in a wheat germ system were inhibited by addition of exogenous signal recognition particle suggesting that interactions of the modified signal peptides with signal recognition particle were normal. The position-dependent effects of the hydrophobic core modifications indicate that structural properties of the core in addition to hydrophobicity are important for signal activity. The parallel effects of the modifications on co-translational translocation and post-translational processing by purified signal peptidase suggest that proteins in the signal peptidase complex might be part of, or intimately associated with, membrane proteins involved in the translocation. A model is proposed in which the NH2-terminal region of the hydrophobic core binds to one subunit of the signal peptidase while the other subunit catalyzes the cleavage.     

Thermodynamics of oligosaccharide binding to a monoclonal antibody specific for a Salmonella O-antigen point to hydrophobic interactions in the binding site.

The thermodynamic characteristics of oligosaccharide binding to an antibody binding site that is dominated by aromatic amino acids suggest that the hydrophobic effect contributes substantially to complex formation as well as hydrogen bonding and van der Waals interactions. A detailed titration microcalorimetric study on the temperature dependence of the binding of a trisaccharide, representing the epitope of a Salmonella O-antigen, showed that its maximum binding to the monoclonal antibody Se155-4 occurs just below room temperature and both enthalpy and entropy changes are strongly dependent on temperature in a mutually compensating manner. The heat capacity change also shows an unusually strong temperature dependence being large and negative above room temperature and positive below. van't Hoff analysis of the temperature dependence of the binding constant yielded a biphasic curve with two apparent intrinsic enthalpy estimations (approximately -100 kJ mol-1 above 18 degrees C and approximately +100 kJ mol-1 below), each very different from the calorimetrically determined enthalpies (ranging from about -60 kJ mol-1 to -20 kJ mol-1). This was interpreted as being due to large enthalpy contributions from concomitant reactions, most notably changes in solvation. Linear plots, -delta H0 versus -T delta S0, observed for temperature-dependent measurements mirror the behavior seen for a series of functional group replacements, suggesting that the molecular and physical origin of these phenomena are closely related and linked to the role of water in complex formation. The thermodynamic results are compared to the mode of binding determined from a 2.05-A resolution structure of the Fab-oligosaccharide complex, and with literature data for the heat capacities of sugars in aqueous solution and for the thermodynamics of carbohydrate binding to transport proteins and lectins.     

Thrombospondin binding to specific sequences within the A alpha- and B beta-chains of fibrinogen

Thrombospondin is a multifunctional adhesive glycoprotein which binds to the surface of resting and activated platelets. Thrombospondin also binds to a variety of proteins, including fibrinogen. The interactions between platelet-bound thrombospondin and fibrinogen are thought to facilitate irreversible platelet aggregation. Both the A alpha- and B beta-chains of fibrinogen specifically bind to thrombospondin. Cyanogen bromide cleavage products of the fibrinogen A alpha- and B beta-chains, and synthetic peptides corresponding to specific regions of these cleavage products were utilized to identify the regions of the fibrinogen A alpha- and B beta-chains which bind to thrombospondin. Cyanogen bromide fragments of the A alpha- and B beta-fibrinogen chains, resolved by gel filtration and reversed-phase chromatography, were examined for thrombospondin binding activity. Thrombospondin specifically bound to the A alpha-chain fragment encompassing residues 92-147 and the B beta-chain fragment encompassing residues 243-305. Analyses of the binding characteristics of two series of overlapping synthetic peptides revealed that peptides corresponding to residues 113-126 of the A alpha-chain and residues 243-252 of the B beta-chain retained thrombospondin binding activity. Separate bovine serum albumin conjugates of the active A alpha-chain and B beta-chain peptides inhibited platelet aggregation. These studies reveal that fibrinogen possesses at least two unique sequences which are recognized by thrombospondin and that such interaction may affect platelet aggregation.     

The surprising complexity of signal sequences

Most secreted and many membrane proteins contain cleavable N-terminal signal sequences that mediate their targeting to and translocation across the endoplasmic reticulum or bacterial cytoplasmic membrane. Recent studies have identified many exceptions to the widely held view that signal sequences are simple, degenerate and interchangeable. Growing evidence indicates that signal sequences contain information that specifies the choice of targeting pathway, the efficiency of translocation, the timing of cleavage and even postcleavage functions. As a consequence, signal sequences can have important roles in modulating protein biogenesis. Based on a synthesis of studies in numerous experimental systems, we propose that substrate-specific sequence elements embedded in a conserved domain structure impart unique and physiologically important functionality to signal sequences.     

A calorimetric study on the binding of six general anesthetics to the hydrophobic core of a model protein

The thermodynamic parameters underlying the binding of six volatile general anesthetics to the hydrophobic core of the four-alpha-helix bundle (Aalpha(2)-L38M)(2) are determined using isothermal titration calorimetry. Chloroform, bromoform, trichloroethylene, benzene, desflurane and fluroxene are shown to bind to the four-alpha-helix bundle with dissociation constants of 880+/-10, 90+/-5, 200+/-10, 900+/-30, 220+/-10 and 790+/-40 microM, respectively. The measured dissociation constants for the binding of the six general anesthetics to the four-alpha-helix bundle (Aalpha(2)-L38M)(2) correlate with their human or animal EC(50) values. The negative enthalpy changes indicate that favorable polar interactions are achieved between bound anesthetic and the adjacent amino acid side chains. Because of its small size and the ability to bind a variety of general anesthetics, the four-alpha-helix bundle (Aalpha(2)-L38M)(2) represents an attractive system for structural studies on anesthetic-protein complexes.     

The specificity of cross-reactivity: promiscuous antibody binding involves specific hydrogen bonds rather than nonspecific hydrophobic stickiness

Proteins are renowned for their specificity of function. There is, however, accumulating evidence that many proteins, from enzymes to antibodies, are functionally promiscuous. Promiscuity is of considerable physiological importance. In the immune system, cross‐reactive or multispecific antibodies are implicated in autoimmune and allergy conditions. In most cases, however, the mechanism behind promiscuity and the relationship between specific and promiscuous activities are unknown. Are the two contradictory? Or can a protein exhibit several unrelated activities each of which is highly specific? To address these questions, we studied a multispecific IgE antibody (SPE7) elicited against a 2,4‐dinitrophenyl hapten (DNP). SPE7 is able to distinguish between closely related derivatives such as NP (nitrophenol) and DNP, yet it can also bind a number of unrelated ligands. We find that, like DNP, the cross‐reactants are themselves bound specifically—close derivatives of these cross‐reactants show very low or no binding to SPE7. It has been suggested that cross‐reactivity is simply due to “hydrophobic stickiness”, nonspecific interactions between hydrophobic ligands and binding sites. However, partitioning experiments reveal that affinity for SPE7 is unrelated to ligand hydrophobicity. These data, combined with crystal structures of SPE7 in complex with four different ligands, demonstrate that each cross‐reactant is bound specifically, forming different hydrogen bonds dependant upon its particular chemistry and the availability of complementary antibody residues. SPE7 is highly homologous to the germline antinitrophenol (NP) antibody B1–8. By comparing the sequences and binding patterns of SPE7 and B1–8, we address the relationship between affinity maturation, specificity, and cross‐reactivity.     

Structures of revertant signal sequences of Escherichia coli ribose binding protein.

Recently we reported (Yi et al., 1994) that the alpha-helical content of the signal peptide of Escherichia coli ribose binding protein, when determined by circular dichroism (CD) and two-dimensional NMR in trifluoroethanol/water solvent, is higher than that of its nonfunctional mutant signal peptide. In the present investigation, the structures of the signal peptides of two revertant ribose binding proteins in the same solvent were also determined with CD and two-dimensional 1H NMR spectroscopy. According to the CD results, both of these revertant signal peptides showed an intermediate helicity between those of wild-type and mutant signal peptides, the helical content of the revertant peptide with higher recovery of the translocation capability being higher. On the other hand, the alpha-helix regions of the wild-type and the revertant peptides as determined by NMR were shown to be the same. This discrepancy may be due to the difference in stability between identical alpha-helical stretches in wild-type and revertant peptides. A good correlation was observed between the helical content of these four ribose binding protein signal peptides in TFE/water as studied by CD and their in vivo translocation activities. It appears, therefore, that both the proper length of the helix and the stability are of functional significance.     

Structural influence of hydrophobic core residues on metal binding and specificity in carbonic anhydrase II

Aromatic residues in the hydrophobic core of human carbonic anhydrase II (CAII) influence metal ion binding in the active site. Residues F93, F95, and W97 are contained in a beta-strand that also contains two zinc ligands, H94 and H96. The aromatic amino acids contribute to the high zinc affinity and slow zinc dissociation rate constant of CAII [Hunt, J. A., and Fierke, C. A. (1997) J. Biol. Chem. 272, 20364-20372]. Substitution of these aromatic amino acids with smaller side chains enhances Cu(2+) affinity while decreasing Co(2+) and Zn(2+) affinity [Hunt, J. A., Mahiuddin, A., & Fierke, C. A. (1999) Biochemistry 38, 9054-9062]. Here, X-ray crystal structures of zinc-bound F93I/F95M/W97V and F93S/F95L/W97M CAIIs reveal the introduction of new cavities in the hydrophobic core, compensatory movements of surrounding side chains, and the incorporation of buried water molecules; nevertheless, the enzyme maintains tetrahedral zinc coordination geometry. However, a conformational change of direct metal ligand H94 as well as indirect (i.e., "second-shell") ligand Q92 accompanies metal release in both F93I/F95M/W97V and F93S/F95L/W97M CAIIs, thereby eliminating preorientation of the histidine ligands with tetrahedral geometry in the apoenzyme. Only one cobalt-bound variant, F93I/F95M/W97V CAII, maintains tetrahedral metal coordination geometry; F93S/F95L/W97M CAII binds Co(2+) with trigonal bipyramidal coordination geometry due to the addition of azide anion to the metal coordination polyhedron. The copper-bound variants exhibit either square pyramidal or trigonal bipyramidal metal coordination geometry due to the addition of a second solvent molecule to the metal coordination polyhedron. The key finding of this work is that aromatic core residues serve as anchors that help to preorient direct and second-shell ligands to optimize zinc binding geometry and destabilize alternative geometries. These geometrical constraints are likely a main determinant of the enhanced zinc/copper specificity of CAII as compared to small molecule chelators.     

Novel DNA binding proteins highly specific to UV-damaged DNA sequences from embryos of Drosophila melanogaster.

Three new proteins which selectively bind to UV-damaged DNA were identified and purified to near homogeneity from UV-irradiated Drosophila melanogaster embryos through several column chromatographies. These proteins, tentatively designated as D-DDB P1, P2 and P3, can be identified as different complex bands in a gel shift assay by using UV-irradiated TC-31 probe DNA. Analysis of the purified D-DDB P1 fraction by native or SDS-polyacrylamide gel electrophoresis and FPLC-Superose 6 gel filtration demonstrated that it is a monomer protein which is a 30 kDa polypeptide. The D-DDB P2 protein is a monopolypeptide with a molecular mass of 14 kDa. Both D-DDB P1 and P2 highly prefer binding to UV-irradiated DNA, and have almost no affinity for non-irradiated DNA. Gel shift assays with either UV-irradiated DNA probes demonstrated that D-DDB P1 may show a preference for binding to (6-4) photoproducts, while D-DDB P2 may prefer binding to pyrimidine dimers. Both these proteins require magnesium ions for binding. D-DDB P1 is an ATP-preferent protein. These findings are discussed in relation to two recently described [Todo and Ryo (1991) Mutat. Res., 273, 85-93; Todo et al. (1993) Nature, 361, 371-374] DNA-binding factors from Drosophila cell extracts. A possible role for these DNA-binding proteins in lesion recognition and DNA-binding proteins in lesion recognition and DNA repair of UV-induced photo-products is discussed.     

A regulatory domain controls the transport activity of a twin-arginine signal peptide

The twin-arginine translocation (Tat) pathway is used by bacteria for the transmembrane transport of folded proteins. Proteins are targeted to the Tat translocase by signal peptides that have common tripartite structures consisting of polar n-regions, hydrophobic h-regions, and polar c-regions. In this work, the signal peptide of [NiFe] hydrogenase-1 from Escherichia coli has been studied. The hydrogenase-1 signal peptide contains an extended n-region that has a conserved primary structure. Genetic and biochemical approaches reveal that the signal peptide n-region is essential for hydrogenase assembly and acts as a regulatory domain controlling transport activity of the signal peptide.