CyclicAMP receptor protein (CRP) regulates transcription of numerous genes in Escherichia coli. Both cAMP and cGMP bind CRP, but only cAMP induces conformational changes that dramatically increase the specific DNA binding activity of the protein. We have shown previously that our protein footprinting technique is sensitive enough to detect conformational changes in CRP by cAMP [Baichoo N, Heyduk T. 1997. Biochemistry 36:10830–10836]. In this work, conformational changes in CRP induced by cAMP and cGMP binding were mapped and quantitatively analyzed by protein footprinting using iron complexed to diethylenetriaminepentaacetic acid ([Fe-DTPA]2−), iron complexed to ethylenediaminediacetic acid ([Fe-EDDA]), iron complexed to desferrioxamine mesylate ([Fe-HDFO]+), and copper complexed to o-phenanthroline ([(OP)2Cu]+) as proteases. These chemical proteases differ in size, charge, and hydrophobicity. Binding of cAMP to CRP resulted in changes in susceptibility to cleavage by all four proteases. Cleavage by [Fe-EDDA] and [Fe-DTPA]2− of CRP-cAMP detected hypersensitivities in the DNA-binding F α-helix, the interdomain hinge, and the ends of the C α-helix, which is involved in intersubunit interactions. [Fe-EDDA] and [Fe-DTPA]2− also detected reductions in cleavage in the D and E α-helices, which are involved in DNA recognition. Cleavage by [Fe-HDFO]+ of CRP-cAMP detected hypersensitivities in β-strand 8, the B α-helix, as well as in parts of the F and C α-helices. [Fe-HDFO]+ also detected protections from cleavage in β-strands 4 to 5 and their intervening loop, β-strand 7, which is part of the nucleotide binding pocket, as well as in the D and E α-helices. Cleavage by [(OP)2Cu]+ of CRP-cAMP detected hypersensitivities in β-strands 9 and 11 as well as in the D and E α-helices. [(OP)2Cu]+ also detected protections in the C α-helix , the interdomain hinge, and β-strands 2–7. Binding of cGMP to CRP resulted in changes in susceptibility to cleavage only by [(OP)2Cu]+, which detected minor protections in β-strands 3–7, the interdomain hinge, and the C α-helix. These results show that binding of cAMP causes structural changes in CRP in the nucleotide binding domain, the interdomain hinge, the DNA binding domain, and regions involved in intersubunit interaction. Structural changes induced by binding of cGMP appear to be very minor and confined to the nucleotide binding domain, the interdomain hinge, and regions involved in intersubunit interaction. Use of different cleaving agents in protein footprinting seems to give a more detailed picture of structural changes than the use of a single protease alone.

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.

Folding of apomyoglobin is characterized by formation of a compact intermediate that contains substantial helicity. To determine whether this intermediate is obligatory or whether the protein can fold directly into the native state via an alternate parallel pathway, we have combined quench-flow hydrogen-exchange pulse labeling techniques with electrospray ionization mass spectrometry. The mass spectra of apomyoglobin obtained at various refolding times suggest that apomyoglobin indeed folds through a single pathway containing an obligatory intermediate with a significant hydrogen-bonded secondary structure content.

We present direct evidence for a change in protein structural specificity due to hydrophobic core packing. High resolution structural analysis of a designed core variant of ubiquitin reveals that the protein is in slow exchange between two conformations. Examination of side-chain rotamers indicates that this dynamic response and the lower stability of the protein are coupled to greater strain and mobility in the core. The results suggest that manipulating the level of side-chain strain may be one way of fine tuning the stability and specificity of proteins.

The tumor suppressor function of the wild-type p53 protein is transdominantly inhibited by tumor-derived mutant p53 proteins. Such transdominant inhibition limits the prospects for gene therapy approaches that aim to introduce wild-type p53 into cancer cells. The molecular mechanism for transdominant inhibition involves sequestration of wild-type p53 subunits into inactive wild-type/mutant hetero-tetramers. Thus, p53 proteins, whose oligomerization specificity is altered so they cannot interact with tumor-derived mutant p53, would escape transdominant inhibition. Aided by the known three-dimensional structure of the p53 tetramerization domain and by trial and error we designed a novel domain with seven amino acid substitutions in the hydrophobic core. A full-length p53 protein bearing this novel domain formed homo-tetramers and had tumor suppressor function, but did not hetero-oligomerize with tumor-derived mutant p53 and resisted transdominant inhibition. Thus, hydrophobic core residues influence the oligomerization specificity of the p53 tetramerization domain.

We have adopted nanoflow electrospray ionization mass spectrometry (ESI-MS) and isothermal titration calorimetry (ITC) to probe the mechanism of peptide recognition by the SH2 domain from the Src family tyrosine kinase protein, Fyn. This domain is involved in the mediation of intracellular signal transduction pathways by interaction with proteins containing phosphorylated tyrosine (Y*) residues. The binding of tyrosyl phosphopeptides can mimic these interactions. Specificity in these interactions has been attributed to the interaction of the Y* and residues proximal and C-terminal to it. Previous studies have established that for specific binding with Fyn, the recognition sequence consists of pTyr-Glu-Glu-Ile. The specific interactions involve the binding of Y* with the ionic, and the Y* + 3 Ile residue with the hydrophobic binding pockets on the surface of the Fyn SH2 domain. In this work, a variation in the Y* + 3 residue of this high-affinity sequence was observed to result in changes in the relative binding affinities as determined in solution (ITC) and in the gas phase (nanoflow ESI-MS). X-ray analysis shows that a feature of the Src family SH2 domains is the involvement of water molecules in the peptide binding site. Under the nanoflow ESI conditions, water molecules appear to be maintained in the Fyn SH2–ligand complex. Compelling evidence for these molecules being incorporated in the SH2–peptide interface is provided by the prevalence of the peaks assigned to water-bound over the water-free complex at high-energy conditions. Thus, the stability of water protein–ligand complex appears to be intimately linked to the presence of water.

An expanded, highly dynamic denatured state of staphylococcal nuclease exhibits a native-like topology in the apparent absence of tight packing and fixed hydrogen bonds (Gillespie JR, Shortle D, 1997, J Mol Biol 268:158–169, 170–184). To address the physical basis of the long-range spatial ordering of this molecule, we probe the effects of perturbations of the sequence and solution conditions on the local chain dynamics of a denatured 101-residue fragment that is missing the first three beta strands. Structural interactions between chain segments are inferred from correlated changes in the motional behavior of residues monitored by 15N NMR relaxation measurements. Restoration of the sequence corresponding to the first three beta strands significantly increases the average order of all chain segments that form the five strand beta barrel including loops but has no effect on the carboxy terminal 30 residues. Addition of the denaturing salt sodium perchlorate enhances ordering over the entire sequence of this fragment. Analysis of seven different substitution mutants points to a complex set of interactions between the hydrophobic segment corresponding to beta strand 5 and the remainder of the chain. General patterns in the data suggest there is a hierarchy of native-like interactions that occur transiently in the denatured state and are consistent with the overall topology of the denatured state ensemble being determined by many coupled local interactions rather than a few highly specific long-range interactions.

Using homology search, structure prediction, and structural characterization methods we show that the C-terminal domains of (1) netrins, (2) complement proteins C3, C4, C5, (3) secreted frizzled-related proteins, and (4) type I procollagen C-proteinase enhancer proteins (PCOLCEs) are homologous with the N-terminal domains of (5) tissue inhibitors of metalloproteinases (TIMPs). The proteins harboring this netrin module (NTR module) fulfill diverse biological roles ranging from axon guidance, regulation of Wnt signaling, to the control of the activity of metalloproteases. With the exception of TIMPs, it is not known at present what role the NTR modules play in these processes. In view of the fact that the NTR modules of TIMPs are involved in the inhibition of matrixin-type metalloproteases and that the NTR module of PCOLCEs is involved in the control of the activity of the astacin-type metalloprotease BMP1, it seems possible that interaction with metzincins could be a shared property of NTR modules and could be critical for the biological roles of the host proteins.

In the past few years, a new generation of fold recognition methods has been developed, in which the classical sequence information is combined with information obtained from secondary structure and, sometimes, accessibility predictions. The results are promising, indicating that this approach may compete with potential-based methods (Rost B et al., 1997, J Mol Biol 270:471–480). Here we present a systematic study of the different factors contributing to the performance of these methods, in particular when applied to the problem of fold recognition of remote homologues.

Our results indicate that secondary structure and accessibility prediction methods have reached an accuracy level where they are not the major factor limiting the accuracy of fold recognition. The pattern degeneracy problem is confirmed as the major source of error of these methods. On the basis of these results, we study three different options to overcome these limitations: normalization schemes, mapping of the coil state into the different zones of the Ramachandran plot, and post-threading graphical analysis.

The biotin carboxyl carrier protein (BCCP) is a subunit of acetyl-CoA carboxylase, a biotin-dependent enzyme that catalyzes the first committed step of fatty acid biosynthesis. In its functional cycle, this protein engages in heterologous protein–protein interactions with three distinct partners, depending on its state of post-translational modification. Apo-BCCP interacts specifically with the biotin holoenzyme synthetase, BirA, which results in the post-translational attachment of biotin to a single lysine residue on BCCP. Holo-BCCP then interacts with the biotin carboxylase subunit of acetyl-CoA carboxylase, which leads to the addition of the carboxylate group of bicarbonate to biotin. Finally, the carboxy-biotinylated form of BCCP interacts with transcarboxylase in the transfer of the carboxylate to acetyl-CoA to form malonyl-CoA. The determinants of protein–protein interaction specificity in this system are unknown. The NMR solution structure of the unbiotinylated form of an 87 residue C-terminal domain fragment (residue 70–156) of BCCP (holoBCCP87) and the crystal structure of the biotinylated form of a C-terminal fragment (residue 77–156) of BCCP from Escherichia coli acetyl-CoA carboxylase have previously been determined. Comparative analysis of these structures provided evidence for small, localized conformational changes in the biotin-binding region upon biotinylation of the protein. These structural changes may be important for regulating specific protein–protein interactions. Since the dynamic properties of proteins are correlated with local structural environments, we have determined the relaxation parameters of the backbone 15N nuclear spins of holoBCCP87, and compared these with the data obtained for the apo protein. The results indicate that upon biotinylation, the inherent mobility of the biotin-binding region and the protruding thumb, with which the biotin group interacts in the holo protein, are significantly reduced.

We carry out a systematic analysis of the correlation between similarity of protein three-dimensional structures and their evolutionary relationships. The structural similarity is quantitatively identified by an all-against-all comparison of the spatial arrangement of secondary structural elements in nonredundant 967 representative proteins, and the evolutionary relationship is judged according to the definition of superfamily in the SCOP database. We find the following symmetry rule: a protein pair that has similar folds but belong to different superfamilies has (with a very rare exception) certain internal symmetry in its common similar folds. Possible reasons behind the symmetry rule are discussed.

We report the solution structure of the chemotactic cytokine (chemokine) vMIP-II. This protein has unique biological activities in that it blocks infection by several different human immunodeficiency virus type 1 (HIV-1) strains. This occurs because vMIP-II binds to a wide range of chemokine receptors, some of which are used by HIV to gain cell entry. vMIP-II is a monomeric protein, unlike most members of the chemokine family, and its structure consists of a disordered N-terminus, followed by a helical turn (Gln25–Leu27), which leads into the first strand of a three-stranded antiparallel β-sheet (Ser29–Thr34; Gly42–Thr47; Gln52–Asp56). Following the sheet is a C-terminal α-helix, which extends from residue Asp60 until Gln68. The final five residues beyond the C-terminal helix (Pro70–Arg74) are in an extended conformation, but several of these C-terminal residues contact the first β-strand. The structure of vMIP-II is compared to other chemokines that also block infection by HIV-1, and the structural basis of its lack of ability to form a dimer is discussed.

The factors that influence the enhanced stability observed experimentally of human rhinovirus 14 (HRV14) upon binding a hydrophobic antiviral drug have been investigated by molecular dynamics. Simulations centered about the HRV14 drug-binding pocket allow the reliable assessment of differences in capsid protein motions of HRV14 and drug-bound HRV14. We propose that the experimentally observed stabilization of the ligated virus arises from higher entropy, rather than enthalpy. Time-averaged interaction energies between the viral protein and molecules occupying the pocket are less favorable in the presence of the drug, consistent with the proposal that the observed stability arises from entropic effects. Interaction energies characterizing subunit–subunit contacts within one viral protomer are found to be substantially stronger than those between two protomers. Such distinction in subunit interaction would have clear implications on assembly and disassembly. Drug binding is found to affect large-scale, collective properties, while leaving local atomic properties unperturbed. Specifically, the simulations reveal a weakening of long-range correlations in atomic motions upon drug binding. On the other hand, neither the fast time scale RMS fluctuations of individual atomic positions nor the fluctuation build-up curves from the capsid β-sandwich forming the drug-binding pocket show a consistent distinction between the drug-bound and drug-free viral simulations. Collectively, the detailed description available from the simulations provides an understanding of the experimental observations on the drug-induced changes in thermal stability and protease sensitivity reported for picornaviruses. The predicted significance of binding entropy can be explored experimentally and should be considered in the design of new antiviral compounds.

Hydrophobic substitutions at solvent-exposed positions in two α-helical regions of the bacteriophage P22 Arc repressor were introduced by combinatorial mutagenesis. In helix A, hydrophobic residues were tolerated individually at each of the five positions examined, but multiple substitutions were poorly tolerated as shown by the finding that mutants with more than two additional hydrophobic residues were biologically inactive. Several inactive helix A variants were purified and found to have reduced thermal stability relative to wild-type Arc, with a rough correlation between the number of polar-to-hydrophobic substitutions and the magnitude of the stability defect. Quite different results were obtained in helix B, where variants with as many as five polar-to-hydrophobic substitutions were found to be biologically active and one variant with three hydrophobic substitutions had a tm 6 °C higher than wild-type. By contrast, a helix A mutant with three similar polar-to-hydrophobic substitutions was 23 °C less stable than wild-type. Also, one set of three polar-to-hydrophobic substitutions in helix B was tolerated when introduced into the wild-type background but not when introduced into an equally active mutant having a nearly identical structure. Context effects occur both when comparing different regions of the same protein and when comparing the same region in two different homologues.

The contribution to the free energy of binding of each of the residues forming the binding site for a human IgG Fc fragment on the surface of the B1 domain of protein G was determined by alanine-scanning mutagenesis. The interface between these two proteins is atypical in that it is smaller than usual, polar in character, and involves two well-defined “knobs-into-holes” interactions. The bulk of the free energy of binding is contributed by three central residues, which make hydrogen bonds across the interface. Of these, the most critical interaction is formed by Glu27, which acts as a charged knob on the surface of the B1 domain, inserting into a polar hole on the Fc fragment. A single alanine mutation of this residue virtually abolishes stable complex formation. Formation of a stable interface between these two proteins is therefore dominated by a small, polar “hot spot.”

Phage-display peptide library analysis of an anti-F actin polyclonal antibody identified 12 amino acid residues of actin that appear, in its X-ray crystal structure, to be grouped together in a surface accessible conformational epitope. Phage epitope mapping was carried out by isolating immune complexes containing members of the J404 nonapeptide phage-display library formed in diluted antiserum and isolated on a protein A affinity matrix. Immunoreactive clones were grown as plaques, replica plated onto nitrocellulose, and labeled with anti-actin immune serum. One hundred and forty-four positively staining clones identified in this way were sequenced. Of these, 54 displayed peptides with sequence similarities. When the most abundantly selected sequence, KQTWQQLWD, was produced as a synthetic peptide and derivatized to ovalbumin, the complex was strongly recognized by the antiserum on Western blots and inhibited the binding of the antibody to immobilized F-actin by 60%. A scrambled version of this sequence WQDKWLQTQ, when coupled to ovalbumin, was not recognized by the antiserum and minimally inhibited binding of antiserum to immobilized F-actin by 10%. KQTWQQLWD contained four residues that corresponded, in frame, to a highly conserved six residue region of the chicken β-actin sequence 351TFQQMW356 (identical residues are shown in bold). Examination of the rabbit skeletal muscle X-ray crystal structure suggested that within a 15 Å radius of W356, nine additional residues were arranged on the actin surface in such a way that they could be mimicked by several of the selected phage sequences with root-mean-square deviation fits of 2.1–2.5 Å. We conclude that phage-display analysis can provide information about the relative location of amino acids on the surfaces of proteins using antibody imprints of the protein surface structure.

Methods for protein structure (3D)–sequence (1D) compatibility evaluation (threading) have been developed during the past decade. The protocol in which a sequence can recognize its compatible structure in the structural library (i.e., the fold recognition or the forward-folding search) is available for the structure prediction of new proteins. However, the reverse protocol, in which a structure recognizes its homologous sequences among a sequence database, named the inverse-folding search, is a more difficult application. In this study, we have investigated the feasibility of the latter approach. A structural library, composed of about 400 well-resolved structures with mutually dissimilar sequences, was prepared, and 163 of them had remote homologs in the library. We examined whether they could correctly seek their homologs by both forward- and inverse-folding searches. The results showed that the inverse-folding protocol is more effective than the forward-folding protocol, once the reference states of the compatibility functions are appropriately adjusted. This adjustment only slightly affects the ability of the forward-folding search. We noticed that the scoring, in which a given sequence is re-mounted onto a structure according to the 3D–1D alignment determined by the dynamic programming method, is only effective in the forward-folding protocol and not in the inverse-folding protocol. Namely, the inverse-folding search works significantly better with the score given by the 3D–1D alignment per se, rather than that obtained by the re-mounting. The implications of these results are discussed.

Penicillin G acylase is an important enzyme in the commercial production of semisynthetic penicillins used to combat bacterial infections. Mutant strains of Providencia rettgeri were generated from wild-type cultures subjected to nutritional selective pressure. One such mutant, Bro1, was able to use 6-bromohexanamide as its sole nitrogen source. Penicillin acylase from the Bro1 strain exhibited an altered substrate specificity consistent with the ability of the mutant to process 6-bromohexanamide. The X-ray structure determination of this enzyme was undertaken to understand its altered specificity and to help in the design of site-directed mutants with desired specificities. In this paper, the structure of the Bro1 penicillin G acylase has been solved at 2.5 Å resolution by molecular replacement. The R-factor after refinement is 0.154 and R-free is 0.165. Of the 758 residues in the Bro1 penicillin acylase heterodimer (α-subunit, 205; β-subunit, 553), all but the eight C-terminal residues of the α-subunit have been modeled based on a partial Bro1 sequence and the complete wild-type P. rettgeri sequence. A tightly bound calcium ion coordinated by one residue from the α-subunit and five residues from the β-subunit has been identified. This enzyme belongs to the superfamily of Ntn hydrolases and uses Oγ of Serβ1 as the characteristic N-terminal nucleophile. A mutation of the wild-type Metα140 to Leu in the Bro1 acylase hydrophobic specificity pocket is evident from the electron density and is consistent with the observed specificity change for Bro1 acylase. The electron density for the N-terminal Gln of the α-subunit is best modeled by the cyclized pyroglutamate form. Examination of aligned penicillin acylase and cephalosporin acylase primary sequences, in conjunction with the P. rettgeri and Escherichia coli penicillin acylase crystal structures, suggests several mutations that could potentially allow penicillin acylase to accept charged β-lactam R-groups and to function as a cephalosporin acylase and thus be used in the manufacture of semi-synthetic cephalosporins.

δ-Crystallin, the major soluble protein component of the avian and reptilian eye lens, is homologous to the urea cycle enzyme argininosuccinate lyase (ASL). In duck lenses there are two δ crystallins, denoted δ1 and δ2. Duck δ2 is both a major structural protein of the lens and also the duck orthologue of ASL, an example of gene recruitment. Although 94% identical to δ2/ASL in the amino acid sequence, δ1 is enzymatically inactive. A series of hybrid proteins have been constructed to assess the role of each structural domain in the enzymatic mechanism. Five chimeras—221, 122, 121, 211, and 112, where the three numbers correspond to the three structural domains and the value of 1 or 2 represents the protein of origin, δ1 or δ2, respectively—were constructed and thermodynamically and kinetically analyzed. The kinetic analysis indicates that only domain 1 is crucial for restoring ASL activity to δ1 crystallin, and that amino acid substitutions in domain 2 may play a role in substrate binding. These results confirm the hypothesis that only one domain, domain 1, is responsible for the loss of catalytic activity in δ1. The thermodynamic characterization of human ASL (hASL) and duck δ1 and δ2 indicate that δ crystallins are slightly less stable than hASL, with the δ1 being the least stable. The ΔGs of unfolding are 57.25, 63.13, and 70.71 kcal mol−1 for δ1, δ2, and hASL, respectively. This result was unexpected, and we speculate that δ crystallins have adapted to their structural role by adopting a slightly less stable conformation that might allow for enhanced protein–protein and protein–solvent interactions.

Binding of the protein Raf to the active form of Ras promotes activation of the MAP kinase signaling pathway, triggering cell growth and differentiation. Raf/Arg89 in the center of the binding interface plays an important role determining Ras–Raf binding affinity. We have investigated experimentally and computationally the Raf-R89K mutation, which abolishes signaling in vivo. The binding to [γ-35S]GTP-Ras of a fusion protein between the Raf-binding domain (RBD) of Raf and GST was reduced at least 175-fold by the mutation, corresponding to a standard binding free energy decrease of at least 3.0 kcal/mol. To compute this free energy and obtain insights into the microscopic interactions favoring binding, we performed alchemical simulations of the RBD, both complexed to Ras and free in solution, in which residue 89 is gradually mutated from Arg into Lys. The simulations give a standard binding free energy decrease of 2.9 ± 1.9 kcal/mol, in agreement with experiment. The use of numerous runs with three different force fields allows insights into the sources of uncertainty in the free energy and its components. The binding decreases partly because of a 7 kcal/mol higher cost to desolvate Lys upon binding, compared to Arg, due to better solvent interactions with the more concentrated Lys charge in the unbound state. This effect is expected to be general, contributing to the lower propensity of Lys to participate in protein–protein interfaces. Large contributions to the free energy change also arise from electrostatic interactions with groups up to 8 Å away, namely residues 37–41 in the conserved effector domain of Ras (including 4 kcal/mol from Ser39 which loses a bifurcated hydrogen bond to Arg89), the conserved Lys84 and Lys87 of Raf, and 2–3 specific water molecules. This analysis will provide insights into the large experimental database of Ras–Raf mutations.