A dissection of specific and non-specific protein-protein interfaces

We compare the geometric and physical-chemical properties of interfaces involved in specific and non-specific protein-protein interactions in crystal structures reported in the Protein Data Bank. Specific interactions are illustrated by 70 protein-protein complexes and by subunit contacts in 122 homodimeric proteins; non-specific interactions are illustrated by 188 pairs of monomeric proteins making crystal-packing contacts selected to bury more than 800 A2 of protein surface. A majority of these pairs have 2-fold symmetry and form "crystal dimers" that cannot be distinguished from real dimers on the basis of the interface size or symmetry. The chemical and amino acid compositions of the large crystal-packing interfaces resemble the protein solvent-accessible surface. These interfaces are less hydrophobic than in homodimers and contain much fewer fully buried atoms. We develop a residue propensity score and a hydrophobic interaction score to assess preferences seen in the chemical and amino acid compositions of the different types of interfaces, and we derive indexes to evaluate the atomic packing, which we find to be less compact at non-specific than at specific interfaces. We test the capacity of these parameters to identify homodimeric proteins in crystal structures, and show that a simple combination of the non-polar interface area and the fraction of buried interface atoms assigns the quaternary structure of 88% of the homodimers and 77% of the monomers in our data set correctly. These success rates increase to 93-95% when the residue propensity score of the interfaces is taken into consideration.     

Structural templates for modeling homodimers

Oligomeric proteins are more abundant in nature than monomeric proteins, and involved in all biological processes. In the absence of an experimental structure, their subunits can be modeled from their sequence like monomeric proteins, but reliable procedures to build the oligomeric assembly are scarce. Template‐based methods, which start from known protein structures, are commonly applied to model subunits. We present a method to model homodimers that relies on a structural alignment of the subunits, and test it on a set of 511 target structures recently released by the Protein Data Bank, taking as templates the earlier released structures of 3108 homodimeric proteins (H‐set), and 2691 monomeric proteins that form dimer‐like assemblies in crystals (M‐set). The structural alignment identifies a H‐set template for 97% of the targets, and in half of the cases, it yields a correct model of the dimer geometry and residue–residue contacts in the target. It also identifies a M‐set template for most of the targets, and some of the crystal dimers are very similar to the target homodimers. The procedure efficiently detects homology at low levels of sequence identities, and points to erroneous quaternary structures in the Protein Data Bank. The high coverage of the target set suggests that the content of the Protein Data Bank already approaches the structural diversity of protein assemblies in nature, and that template‐based methods should become the choice method for modeling oligomeric as well as monomeric proteins.     

Peptide segments in protein-protein interfaces

An important component of functional genomics involves the understanding of protein association. The interfaces resulting from protein-protein interactions - (i) specific, as represented by the homodimeric quaternary structures and the complexes formed by two independently occurring protein components, and (ii) non-specific, as observed in the crystal lattice of monomeric proteins - have been analysed on the basis of the length and the number of peptide segments. In 1000 A2 of the interface area, contributed by a polypeptide chain, there would be 3.4 segments in homodimers, 5.6 in complexes and 6.3 in crystal contacts. Concomitantly, the segments are the longest (with 8.7 interface residues) in homodimers. Core segments (likely to contribute more towards binding) are more in number in homodimers (1.7) than in crystal contacts (0.5), and this number can be used as one of the parameters to distinguish between the two types of interfaces. Dominant segments involved in specific interactions, along with their secondary structural features, are enumerated.     

Versatile architecture of a bacterial aconitase B and its catalytic performance in the sequential reaction coupled with isocitrate dehydrogenase

Aconitase B (AcnB) and isocitrate dehydrogenase (ICDH) catalyze the sequential reaction in the Krebs cycle. Since each enzyme was characterized as an independent protein in a diluted condition, the catalytic performance within the cellular metabolism remains unclear. In particular, high macromolecular concentration in the cytosol promotes weak interactions, which affects structure and function of the proteins. We found that the two bacterial enzymes exhibit variable catalytic performance of the sequential reaction, depending on the oligomerization state. The small-angle solution X-ray scattering and the chemical crosslinking analyses revealed that not only the two enzymes but also the fusion protein of the two enzymes assume homodimers in solution. Interestingly, the fusion protein maintains the homodimeric architecture of ICDH, but not AcnB. Instead, one of the two monomeric AcnB regions associates with the homodimeric ICDH region. The fusion protein displayed different catalytic performance of the sequential reaction from that observed in the mixture of the AcnB and ICDH proteins in an equimolar ratio. Connecting the two proteins by a flexible linker yielded a locally high concentration to promote the weak protein-protein interaction. The versatile architecture of AcnB may alter the metabolic process involving the Krebs cycle.     

Synthetic symmetrization in the crystallization and structure determination of CelA from Thermotoga maritima

Protein crystallization continues to be a major bottleneck in X‐ray crystallography. Previous studies suggest that symmetric proteins, such as homodimers, might crystallize more readily than monomeric proteins or asymmetric complexes. Proteins that are naturally monomeric can be made homodimeric artificially. Our approach is to create homodimeric proteins by introducing single cysteines into the protein of interest, which are then oxidized to form a disulfide bond between the two monomers. By introducing the single cysteine at different sequence positions, one can produce a variety of synthetically dimerized versions of a protein, with each construct expected to exhibit its own crystallization behavior. In earlier work, we demonstrated the potential utility of the approach using T4 lysozyme as a model system. Here we report the successful application of the method to Thermotoga maritima CelA, a thermophilic endoglucanase enzyme with low sequence identity to proteins with structures previously reported in the Protein Data Bank. This protein had resisted crystallization in its natural monomeric form, despite a broad survey of crystallization conditions. The synthetic dimerization of the CelA mutant D188C yielded well‐diffracting crystals with molecules in a packing arrangement that would not have occurred with native, monomeric CelA. A 2.4 Å crystal structure was determined by single anomalous dispersion using a seleno‐methionine derivatized protein. The results support the notion that synthetic symmetrization can be a useful approach for enlarging the search space for crystallizing monomeric proteins or asymmetric complexes.     

On the significance of Toc-GTPase homodimers

Precursor protein translocation across the outer chloroplast membrane depends on the action of the Toc complex, containing GTPases as recognizing receptor components. The G domains of the GTPases are known to dimerize. In the dimeric conformation an arginine contacts the phosphate moieties of bound nucleotide in trans. Kinetic studies suggested that the arginine in itself does not act as an arginine finger of a reciprocal GTPase-activating protein (GAP). Here we investigate the specific function of the residue in two GTPase homologues. Arginine to alanine replacement variants have significantly reduced affinities for dimerization compared with wild-type GTPases. The amino acid exchange does not impact on the overall fold and nucleotide binding, as seen in the monomeric x-ray crystallographic structure of the Arabidopsis Toc33 arginine-alanine replacement variant at 2.0A. We probed the catalytic center with the transition state analogue GDP/AlF(x) using NMR and analytical ultracentrifugation. AlF(x) binding depends on the arginine, suggesting the residue can play a role in catalysis despite the non-GAP nature of the homodimer. Two non-exclusive functional models are discussed: 1) the coGAP hypothesis, in which an additional factor activates the GTPase in homodimeric form; and 2) the switch hypothesis, in which a protein, presumably the large Toc159 GTPase, exchanges with one of the homodimeric subunits, leading to activation.     

NMR analysis of the monomeric form of a mutant unliganded bovine neurophysin: comparison with the crystal structure of a neurophysin dimer

Determination of the structure of the unliganded monomeric state of neurophysin is central to an understanding of the allosteric relationship between neurophysin peptide-binding and dimerization. We examined this state by NMR, using the weakly dimerizing H80E mutant of bovine neurophysin-I. The derived structure, to which more than one conformer appeared to contribute, was compared with the crystal structure of the unliganded des 1-6 bovine neurophysin-II dimer. Significant conformational differences between the two proteins were evident in the orientation of the 3,10 helix, in the 50-58 loop, in beta-turns, and in specific intrachain contacts between amino- and carboxyl domains. However, both had similar secondary structures, in independent confirmation of earlier circular dichroism studies. Previously suggested interactions between the amino terminus and the 50-58 loop in the monomer were also confirmed. Comparison of the observed differences between the two proteins with demonstrated effects of dimerization on the NMR spectrum of bovine neurophysin-I, and preliminary investigation of the effects of dimerization on H80E spectra, allowed tentative distinction between the contributions of sequence and self-association differences to the difference in conformation. Regions altered by dimerization encompass most binding site residues, providing a potential explanation of differences in binding affinity between the unliganded monomeric and dimeric states. Differences between monomer and dimer states in turns, interdomain contacts, and within the interdomain segment of the 50-58 loop suggest that the effects of dimerization on intrasubunit conformation reflect the need to adjust the relative positions of the interface segments of the two domains for optimal interaction with the adjacent subunit and/or reflect the dual role of some residues as participants both at the interface and in interdomain contacts.     

A comparison of the folding of two knotted proteins: YbeA and YibK

The extraordinary topology of proteins belonging to the alpha/beta-knot superfamily of proteins is unexpected, due to the apparent complexities involved in the formation of a deep trefoil knot in a polypeptide backbone. Despite this, an increasing number of knotted structures are being identified; how such proteins fold remains a mystery. Studies on the dimeric protein YibK from Haemophilus influenzae have led to the characterisation of its folding pathway in some detail. To complement research into the folding of YibK, and to address whether folding pathways are conserved for members of the alpha/beta-knot superfamily, the structurally similar knotted protein YbeA from Escherichia coli has been studied. A comprehensive thermodynamic and kinetic analysis of the folding of YbeA is presented here, and compared to that of YibK. Both fold via an intermediate state populated under equilibrium conditions that is monomeric and considerably structured. The unfolding/refolding kinetics of YbeA are simpler than those found for YibK and involve two phases attributed to the formation of a monomeric intermediate state and a dimerisation step. In contrast to YibK, a change in the rate-determining step on the unfolding pathway for YbeA is observed with a changing concentration of urea. Despite this difference, both proteins fold by a mechanism involving at least one sequential monomeric intermediate that has properties similar to that observed during the equilibrium unfolding. The rate of dimerisation observed for YbeA and YibK is very similar, as is the rate constant for formation of the kinetic monomeric intermediate that precedes dimerisation. The findings suggest that relatively slow folding and dimerisation may be common attributes of knotted proteins.     

Folding kinetics of the S100A11 protein dimer studied by time-resolved electrospray mass spectrometry and pulsed hydrogen-deuterium exchange

This study reports the application of electrospray ionization (ESI) mass spectrometry (MS) with on-line rapid mixing for millisecond time-resolved studies of the refolding and assembly of a dimeric protein complex. Acid denaturation of S100A11 disrupts the native homodimeric protein structure. Circular dichroism and HSQC nuclear magnetic resonance measurements reveal that the monomeric subunits unfold to a moderate degree but retain a significant helicity and some tertiary structural elements. Following a rapid change in solution conditions to a slightly basic pH, the native protein reassembles with an effective rate constant of 6 s(-)(1). The ESI charge state distributions measured during the reaction suggest the presence of three kinetic species, namely, a relatively unfolded monomer (M(U)), a more tightly folded monomeric reaction intermediate (M(F)), and dimeric S100A11. These three forms exhibit distinct calcium binding properties, with very low metal loading levels for M(U), up to two calcium ions for M(F), and up to four for the dimer. Surprisingly, on-line pulsed hydrogen-deuterium exchange (HDX) reveals that each of the monomeric forms of the protein comprises two subspecies that can be distinguished on the basis of their isotope exchange levels. As the reaction proceeds, the more extensively labeled species are depleted. The exponential nature of the measured intensity-time profiles implies that the rate-determining step of the overall process is a unimolecular event. The kinetics are consistent with a sequential folding and assembly mechanism involving two increasingly nativelike monomeric intermediates en route to the native S100A11 dimer.     

Thermodynamic and kinetic characterization of the association of triosephosphate isomerase: the role of diffusion

It is known that diffusion plays a central role in the folding of small monomeric proteins and in the rigid-body association of proteins, however, the role of diffusion in the association of the folding intermediates of oligomeric proteins has been scarcely explored. In this work, catalytic activity and fluorescence measurements were used to study the effect of viscosity in the unfolding and refolding of the homodimeric enzyme triosephosphate isomerase from Saccharamyces cerevisiae. Two transitions were found by equilibrium and kinetic experiments, suggesting a three-state model with a monomeric intermediate. Glycerol barely affects DeltaG(0)(fold) whereas DeltaG(0)(assoc) becomes more favourable in the presence of the cosolvent. From 0 to 60% (v/v) glycerol, the association rate constant showed a near unitary dependence on solvent viscosity. However, at higher glycerol concentrations deviations from Kramers theory were observed. The dissociation rate constant showed a viscosity effect much higher than one. This may be related to secondary effects such as short-range glycerol-induced repulsion between monomers. Nevertheless, after comparison under isostability conditions, a slope near one was also observed for the dissociation rate. These results strongly suggest that the bimolecular association producing the native dimer is limited by diffusional events of the polypeptide chains through the solvent.     

Elucidation of the disulfide bridge pattern of the recombinant human growth and differentiation factor 5 dimer and the interchain Cys/Ala mutant monomer

Growth and differentiation factor 5 (GDF5) is involved in many developmental processes such as chondrogenesis and joint and bone formation. A recombinant monomeric human GDF5 mutant rGDF5(C84A) is in vitro as potent as the dimeric native form, and clinical investigations of rGDF5(C84A) are in progress. Native homodimeric GDF5 belongs to the transforming growth factor β (TGF-β) superfamily; each monomer contains a cystine knot formed by three intrachain disulfide bridges, and the monomers are connected via an interchain disulfide bridge. The disulfide bridge pattern of recombinant homodimeric rGDF5 was recently elucidated by X-ray diffraction. A combination of proteolytic degradation with thermolysin, separation of the generated fragments by reverse-phase high-performance liquid chromatography (RP–HPLC), and subsequent analyses of the disulfide-linked peptides by electrospray–mass spectrometry and matrix-assisted laser desorption/ionization time-of-flight (MALDI–TOF) mass spectrometry, amino acid analysis, and Edman degradation led to the unambiguous identification of the disulfide bridge pattern of the monomeric mutant rGDF5(C84A) and of the homodimeric rGDF5 in solution. The cystine knot of homodimeric rGDF5 exhibits the pattern Cys1-Cys5, Cys2-Cys6, and Cys3-Cys7 (three intrachain disulfide bonds), and the monomers are connected by a single interchain disulfide bridge (Cys4-Cys4) in accordance with other members of the TGF-β superfamily. The monomeric mutant rGDF5(C84A) exhibits the same cystine knot pattern as homodimeric rGDF5.     

Interresidue contacts in proteins and protein-protein interfaces and their use in characterizing the homodimeric interface

The environment of amino acid residues in protein tertiary structures and three types of interfaces formed by protein-protein association--in complexes, homodimers, and crystal lattices of monomeric proteins--has been analyzed in terms of the propensity values of the 20 amino acid residues to be in contact with a given residue. On the basis of the similarity of the environment, twenty residues can be divided into nine classes, which may correspond to a set of reduced amino acid alphabet. There is no appreciable change in the environment in going from the tertiary structure to the interface, those participating in the crystal contacts showing the maximum deviation. Contacts between identical residues are very prominent in homodimers and crystal dimers and arise due to 2-fold related association of residues lining the axis of rotation. These two types of interfaces, representing specific and nonspecific associations, are characterized by the types of residues that partake in "self-contacts"--most notably Leu in the former and Glu in the latter. The relative preference of residues to be involved in "self-contacts" can be used to develop a scoring function to identify homodimeric proteins from crystal structures. Thirty-four percent of such residues are fully conserved among homologous proteins in the homodimer dataset, as opposed to only 20% in crystal dimers. Results point to Leu being the stickiest of all amino acid residues, hence its widespread use in motifs, such as leucine zippers.     

The dimeric dihydroorotate dehydrogenase A from Lactococcus lactis dissociates reversibly into inactive monomers

The flavoenzyme dihydroorotate dehydrogenase A from Lactococcus lactis is a homodimeric protein of 311 residues/subunit, and the two active sites are positioned at a distance from the dimer interface. To promote formation of the monomeric form of the enzyme, we changed the residues involved in formation of two salt bridges formed between the residues Glu206 of the one polypeptide and Lys296 of the other polypeptide. The mutant enzymes formed inactive precipitates when cells were grown at 37 degrees C, but remained soluble and active when cells were grown at 25 degrees C. The salt bridges were not needed for activity, because the mutant enzymes in which one of the residues was converted to an alanine (E206A or K296A) retained almost full activity. The mutant enzymes in which the charge of one of the residues of the salt bridge was inverted (i.e., E206K or K296E) were severely impaired. The double mutant E206K-K296E, which has the possibility of forming salt bridges in the opposite orientation of the wild type, was fully active in concentrated solutions, but dissociated into inactive monomers upon dilution. The K(D) for the dimer to monomer dissociation reaction was 12 microM, and dimer formation was favored by the product, orotate, or by high ionic strength, indicating that the hydrophobic interactions are important for the subunit contacts. Wild-type dihydroorotate dehydrogenase A was similarly found to dissociate into inactive monomers, but with a K(D) for dissociation equal to 0.12 microM. These results imply that the dimeric state is necessary for activity of the enzyme.     

From monomeric to homodimeric endonucleases and back: engineering novel specificity of LAGLIDADG enzymes

Monomeric homing endonucleases of the LAGLIDADG family recognize DNA in a bipartite manner, reflecting the underlying structural assembly of two protein domains (A and B) related by pseudo 2-fold symmetry. This architecture allows for changes in DNA specificity via the distinct combination of these half-site domains. The key to engineering such hybrid proteins lies in the LAGLIDADG two-helix bundle that forms both the domain interface and the endonuclease active site. In this study, we utilize domain A of the monomeric I-DmoI to demonstrate the feasibility of generating functional homodimeric endonucleases that recognize palindromic DNA sequences derived from the original, non-palindromic target. Wild-type I-DmoI domain A is capable of forming a homodimer (H-DmoA) that binds tightly to, but does not cleave efficiently, its anticipated DNA target. Partial restoration of DNA cleavage ability was obtained by re-engineering the LAGLIDADG dimerization interface (H-DmoC). Upon fusing two copies of H-DmoC via a short peptide linker, a novel, site-specific DNA endonuclease was created (H-DmoC2). Like I-DmoI, H-DmoC2 is thermostable and cleaves the new target DNA to generate the predicted 4 nt 3'-OH overhangs but, unlike I-DmoI, H-DmoC2 retains stringent cleavage specificity when substituting Mn2+ for Mg2+ as co-factor. This novel endonuclease allows speculation regarding specificity of monomeric LAGLIDADG proteins, while it supports the evolutionary genesis of these proteins by a gene duplication event.     

Energetics of MazG Unfolding in Correlation with Its Structural Features

MazG is a homodimeric α-helical protein that belongs to the superfamily of all-α NTP pyrophosphatases. Its function has been connected to the regulation of the toxin-antitoxin module mazEF, implicated in programmed growth arrest/cell death of Escherichia coli cells under conditions of amino acid starvation. The goal of the first detailed biophysical study of a member of the all-α NTP pyrophosphatase superfamily, presented here, is to improve molecular understanding of the unfolding of this type of proteins. Thermal unfolding of MazG monitored by differential scanning calorimetry, circular dichroism spectroscopy, and fluorimetry at neutral pH in the presence of a reducing agent (dithiothreitol) can be successfully described as a reversible four-state transition between a dimeric native state, two dimeric intermediate states, and a monomeric denatured state. The first intermediate state appears to have a structure similar to that of the native state while the final thermally denatured monomeric state is not fully unfolded and contains a significant fraction of residual α-helical structure. In the absence of dithiothreitol, disulfide cross-linking causes misfolding of MazG that appears to be responsible for the formation of multimeric aggregates. MazG is most stable at pH 7-8, while at pH < 6, it exists in a molten-globule-like state. The thermodynamic parameters characterizing each step of MazG denaturation transition obtained by global fitting of the four-state model to differential scanning calorimetry, circular dichroism, and fluorimetry temperature profiles are in agreement with the observed structural characteristics of the MazG conformational states and their assumed functional role.     

Structural characterization of human S100A16, a low-affinity calcium binder

The homodimeric structure of human S100A16 in the apo state has been obtained both in the solid state and in solution, resulting in good agreement between the structures with the exception of two loop regions. The homodimeric solution structure of human S100A16 was also calculated in the calcium(II)-bound form. Differently from most S100 proteins, the conformational rearrangement upon calcium binding is minor. This characteristic is likely to be related to the weak binding affinity of the protein for the calcium(II) ions. In turn, this is ascribed to the lack of the glutamate residue at the end of the S100-specific N-domain binding site, which in most S100 proteins provides two important side chain oxygen atoms as calcium(II) ligands. Furthermore, the presence of hydrophobic interactions stronger than for other S100 proteins, present in the closed form of S100A16 between the third and fourth helices, likely make the closed structure of the second EF-hand particularly stable, so even upon calcium(II) binding such a conformation is not disrupted.     

Photosystem II: evolutionary perspectives

Based on the current model of its structure and function, photosystem II (PSII) seems to have evolved from an ancestor that was homodimeric in terms of its protein core and contained a special pair of chlorophylls as the photo-oxidizable cofactor. It is proposed that the key event in the evolution of PSII was a mutation that resulted in the separation of the two pigments that made up the special chlorophyll pair, making them into two chlorophylls that were neither special nor paired. These ordinary chlorophylls, along with the two adjacent monomeric chlorophylls, were very oxidizing: a property proposed to be intrinsic to monomeric chlorophylls in the environment provided by reaction centre (RC) proteins. It seems likely that other (mainly electrostatic) changes in the environments of the pigments probably tuned their redox potentials further but these changes would have been minor compared with the redox jump imposed by splitting of the special pair. This sudden increase in redox potential allowed the development of oxygen evolution. The highly oxidizing homodimeric RC would probably have been not only inefficient in terms of photochemistry and charge storage but also wasteful in terms of protein or pigments undergoing damage due to the oxidative chemistry. These problems would have constituted selective pressures in favour of the lop-sided, heterodimeric system that exists as PSII today, in which the highly oxidized species are limited to only one side of the heterodimer: the sacrificial, rapidly turned-over D1 protein. It is also suggested that one reason for maintaining an oxidizable tyrosine, TyrD, on the D2 side of the RC, is that the proton associated with its tyrosyl radical, has an electrostatic role in confining P(+) to the expendable D1 side.     

The bacterial ATPase SecA functions as a monomer in protein translocation

The ATPase SecA drives the post-translational translocation of proteins through the SecY channel in the bacterial inner membrane. SecA is a dimer that can dissociate into monomers under certain conditions. To address the functional importance of the monomeric state, we generated an Escherichia coli SecA mutant that is almost completely monomeric (>99%), consistent with predictions from the crystal structure of Bacillus subtilis SecA. In vitro, the monomeric derivative retained significant activity in various assays, and in vivo, it sustained 85% of the growth rate of wild type cells and reduced the accumulation of precursor proteins in the cytoplasm. Disulfide cross-linking in intact cells showed that mutant SecA is monomeric and that even its parental dimeric form is dissociated. Our results suggest that SecA functions as a monomer during protein translocation in vivo.     

Drosophila Calpain B is monomeric and autolyzes intramolecularly

Drosophila calpains, Calpain A and Calpain B, show typical calpain domain structures similar to mammalian calpains. However, the small subunit of mammalian calpains, shown to be essential in both genetic and biochemical aspects, is absent in Drosophila calpains and is not required for enzymatic activity. How they compensate for the lack of small subunit is mostly unknown. Here we conducted experiments using recombinant Drosophila Calpain B for further characterization of the enzyme with particular focuses on two issues: possibility of homodimerization and mode of autolysis. The native molecular weight of Calpain B indicates that the active enzyme is primarily monomeric. Co-expression of two recombinant Calpain B proteins each with a unique affinity tag and a subsequent single round of affinity tag purification resulted in isolation of only one recombinant calpain type, suggesting there is no homodimeric interaction. Also the C-termini of Drosophila calpains lack many of the key hydrophobic residues considered to be important in the dimerization of mammalian calpains. Further, initial autolysis of Calpain B seems to occur intramolecularly, which supports the monomeric nature of Drosophila calpains. These results strongly suggest that dimerization is not an essential requirement for Drosophila calpains.