Energetics and kinetics of cooperative cofilin-actin filament interactions

We have evaluated the thermodynamic parameters associated with cooperative cofilin binding to actin filaments, accounting for contributions of ion-linked equilibria, and determined the kinetic basis of cooperative cofilin binding. Ions weaken non-contiguous (isolated, non-cooperative) cofilin binding to an actin filament without affecting cooperative filament interactions. Non-contiguous cofilin binding is coupled to the dissociation of approximately 1.7 thermodynamically bound counterions. Counterion dissociation contributes approximately 40% of the total cofilin binding free energy (in the presence of 50 mM KCl). The non-contiguous and cooperative binding free energies are driven entirely by large, positive entropy changes, consistent with a cofilin-mediated increase in actin filament structural dynamics. The rate constant for cofilin binding to an isolated site on an actin filament is slow and likely to be limited by filament breathing. Cooperative cofilin binding arises from an approximately tenfold more rapid association rate constant and an approximately twofold slower dissociation rate constant. The more rapid association rate constant is presumably a consequence of cofilin-dependent changes in the average orientation of subdomain 2, subunit angular disorder and filament twist, which increase the accessibility of a neighboring cofilin-binding site on an actin filament. Cooperative association is more rapid than binding to an isolated site, but still slow for a second-order reaction, suggesting that cooperative binding is limited also by binding site accessibility. We suggest that the dissociation of actin-associated ions weakens intersubunit interactions in the actin filament lattice that enhance cofilin-binding site accessibility, favor cooperative binding and promote filament severing.     

Determination of binding constants for cooperative site-specific protein-DNA interactions using the gel mobility-shift assay

We have investigated the question of whether the gel mobility-shift assay can provide data that are useful to the demonstration of cooperativity in the site-specific binding of proteins to DNA. Three common patterns of protein-DNA interaction were considered: (i) the cooperative binding of a protein to two sites (illustrated by the Escherichia coli Gal repressor); (ii) the cooperative binding of a bidentate protein to two sites (illustrated by the E. coli Lac repressor); and (iii) the cooperative binding of a protein to three sites (illustrated by the lambda cI repressor). A simple, rigorous, and easily extendable statistical mechanical approach to the derivation of the binding equations for the different patterns is presented. Both simulated and experimental data for each case are analyzed. The mobility-shift assay provides estimates of the macroscopic binding constants for each step of ligation based on its separation of liganded species by the number of ligands bound. Resolution of the binding constants depends on the precision with which the equilibrium distribution of liganded species is determined over the entire range of titration of each of the sites. However, the evaluation of cooperativity from the macroscopic binding constants is meaningful only for data that are also accurate. Some criteria that are useful in evaluating accuracy are introduced and illustrated. Resolution of cooperative effects is robust only for the simplest case, in which there are two identical protein binding sites. In this case, cooperative effects of up to 1,000-fold are precisely determined. For heterogeneous sites, cooperative effects of greater than 1,000-fold are resolvable, but weak cooperativity is masked by the heterogeneity. For three-site systems, only averaged pair-wise cooperative effects are resolvable.     

Gene 5 protein-DNA complex: modeling binding interactions

A helical (not toroidal) complex consisting of eight gene 5 protein dimers per turn is proposed for the extension of DNA from dimer to dimer using known bond length constraints, postulated protein-nucleic acid interactions (determined from NMR and chemical modification studies), other physical properties of the complex, and data from electron micrographs. The binding channel has been dictated by these known parameters and the relative ease of geometrically fitting these constituents. This channel is different from that previously reported by other modelers. The channel lies underneath the long arm "claw-like" extension of the monomer, so that it rests inside the outer surface of the protein complex. An explanation is proposed for the two binding modes, n = 4 (the predominate mode) and n = 3, based on the weak binding interaction of Tyrosine 34. Also, the site of the less mobile nucleic acid base as reported from ESR studies (S.-C. Kao, E.V. Bobst, G.T. Pauly and A.M. Bobst, J. Biom. Struc. Dyn. 3,261 (1985)) is postulated as involving the fourth nucleotide, and this particular base is stacked between Tyrosine 34 and Phenylalanine 73'.     

Energetics of sequence-specific protein-DNA association: binding of integrase Tn916 to its target DNA

The DNA binding domain of the transposon Tn916 integrase (INT-DBD) binds to its DNA target site by positioning the face of a three-stranded antiparallel beta-sheet within the major groove. Binding of INT-DBD to a 13 base pair duplex DNA target site was studied by isothermal titration calorimetry, differential scanning calorimetry, thermal melting followed by circular dichroism spectroscopy, and fluorescence spectroscopy. The observed heat capacity change accompanying the association reaction (DeltaC(p)) is temperature-dependent, decreasing from -1.4 kJ K(-1) mol(-1) at 4 degrees C to -2.9 kJ K(-1) mol(-1) at 30 degrees C. The reason is that the partial molar heat capacities of the free protein, the free DNA duplex, and the protein-DNA complex are not changing in parallel when the temperature increases and that thermal motions of the protein and the DNA are restricted in the complex. After correction for this effect, DeltaC(p) is -1.8 kJ K(-1) mol(-1) and temperature-independent. However, this value is still higher than DeltaC(p) of -1.2 kJ K(-1) mol(-1) estimated by semiempirical methods from dehydration of surface area buried at the complex interface. We propose that the discrepancy between the measured and the structure-based prediction of binding energetics is caused by incomplete dehydration of polar groups in the complex. In support, we identify cavities at the interface that are large enough to accommodate approximately 10 water molecules. Our results highlight the difficulties of structure-based prediction of DeltaC(p) (and other thermodynamic parameters) and emphasize how important it is to consider changes of thermal motions and soft vibrational modi in protein-DNA association reactions. This requires not only a detailed investigation of the energetics of the complex but also of the folding thermodynamics of the protein and the DNA alone, which are described in the accompanying paper [Milev et al. (2003) Biochemistry 42, 3492-3502].     

Non-specific protein-DNA interactions control I-CreI target binding and cleavage

Homing endonucleases represent protein scaffolds that provide powerful tools for genome manipulation, as these enzymes possess a very low frequency of DNA cleavage in eukaryotic genomes due to their high specificity. The basis of protein-DNA recognition must be understood to generate tailored enzymes that target the DNA at sites of interest. Protein-DNA interaction engineering of homing endonucleases has demonstrated the potential of these approaches to create new specific instruments to target genes for inactivation or repair. Protein-DNA interface studies have been focused mostly on specific contacts between amino acid side chains and bases to redesign the binding interface. However, it has been shown that 4 bp in the central DNA sequence of the 22-bp substrate of a homing endonuclease (I-CreI), which do not show specific protein-DNA interactions, is not devoid of content information. Here, we analyze the mechanism of target discrimination in this substrate region by the I-CreI protein, determining how it can occur independently of the specific protein-DNA interactions. Our data suggest the important role of indirect readout in this substrate region, opening the possibility for a fully rational search of new target sequences, thus improving the development of redesigned enzymes for therapeutic and biotechnological applications.     

Energetics of sequence-specific protein-DNA association: computational analysis of integrase Tn916 binding to its target DNA

The N-terminal domain of the bacterial integrase Tn916 specifically recognizes the 11 bp DNA target site by positioning the face of a three-stranded beta-sheet into the major groove. Binding is linked to structural adaptation. We have characterized INT-DBD binding to DNA in detail by calorimetry [Milev, S., Gorfe, A., Karshikoff, A., Clubb, R. T., Bosshard, H. R., and Jelesarov, I. (2003) Biochemistry 42, 3481-3491]. Our thermodynamic analysis has indicated that the major driving force of association is the hydrophobic effect while polar interactions contribute less. To gain more comprehensive information about the binding process, we performed a computational analysis of the binding free energy and report here the results. A hybrid molecular mechanics/continuum approach was followed. The total binding free energy is predicted with reasonable accuracy. The calculations confirm that nonpolar effects stabilize the protein-DNA complex while electrostatics opposes binding. Structural changes optimizing surface complementarity are costly in terms of energy. The energetic consequences from the replacement of nine DNA-contacting residues by alanine were investigated. The calculations correctly predict the binding affinity decrease of eight mutations and the destabilizing effect of one wild-type residue. Bulky side chains stabilize the wild-type complex through packing interactions and favorable nonpolar dehydration, but the net nonpolar energy changes do not correlate with the relative affinity loss upon mutation. Discrete protein-DNA electrostatic interactions may be net stabilizing or net destabilizing depending on the local environment. In contrast to nonpolar energy changes, the magnitude of the electrostatic free energy ranks the mutations according to the experimentally measured DeltaDeltaG. Free energy decomposition analysis from a structural perspective leads to detailed information about the thermodynamic strategy used by INT-DBD for sequence-specific DNA binding.     

Cooperative protein-DNA interactions: effects of KCl on lambda cI binding to OR.

The effects of monovalent salt activity on the site-specific and cooperative interactions of cI repressor with its three operator sites OR were studied by using quantitative DNase I footprint titration methods. Individual-site binding isotherms were obtained for binding repressor dimers to each site of wild-type OR and to mutant operator templates in which binding to one or two sites has been eliminated. The standard Gibbs energies for intrinsic binding, delta G1, delta G2, and delta G3, and cooperative interactions, delta G12 and delta G23, were determined at each condition (range 50-200 mM KCl). It is found that the dimer affinity for each of the three sites increases as [KCl] decreases, a striking result given that the monomer-dimer equilibrium shifts toward monomer formation under identical solution conditions [Koblan, K. S., & Ackers, G. K. (1991) Biochemistry (preceding paper in this issue)]. The magnitudes of ion-linked effects are found to differ at the three operator sites, while the intrinsic interaction binding free energies for sites OR1 and OR3 change in parallel over the entire range of [KCl]. The KCl dependencies at OR1 and OR3 represent the average release of 3.7 +/- 0.6 and 3.8 +/- 0.6 apparent ions, respectively. By contrast, the KCl dependency of OR2 binding corresponds to the displacement of 5.2 +/- 0.7 apparent ions. The ability of cI repressor to discriminate between the three operator sites thus appears linked to ion binding/release reactions.     

An analysis of the relationship between hydration and protein-DNA interactions.

Eleven protein-DNA crystal structures were analyzed to test the hypothesis that hydration sites predicted in the first hydration shell of DNA mark the positions where protein residues hydrogen-bond to DNA. For nine of those structures, protein atoms, which form hydrogen bonds to DNA bases, were found within 1.5 A of the predicted hydration positions in 86% of the interactions. The correspondence of the predicted hydration sites with the hydrogen-bonded protein side chains was significantly higher for bases inside the conserved DNA recognition sequences than outside those regions. In two CAP-DNA complexes, predicted base hydration sites correctly marked 71% (within 1.5 A) of protein atoms, which form hydrogen bonds to DNA bases. Phosphate hydration was compared to actual protein binding sites in one CAP-DNA complex with 78% marked contacts within 2.0 A. These data suggest that hydration sites mark the binding sites at protein-DNA interfaces.     

The mechanism of RNA binding to TRAP: initiation and cooperative interactions

The trp RNA-binding Attenuation Protein (TRAP) from Bacillus subtilis is an 11-subunit protein that binds a series of 11 GAG and UAG repeats separated by two to three-spacer nucleosides in trp leader mRNA. The structure of TRAP bound to an RNA containing 11 GAG repeats shows that the RNA wraps around the outside of the protein ring with each GAG interacting with the protein in nearly identical fashion. The only direct hydrogen bond interactions between the protein and the RNA backbone are to the 2'-hydroxyl groups on the third G of each repeat. Replacing all 11 of these guanosines with deoxyriboguanosine eliminates measurable binding to TRAP. In contrast, a single riboguanosine in an otherwise entirely DNA oligonucleotide dramatically stabilizes TRAP binding, and facilitates the interaction of the remaining all-DNA portion with the protein. Studies of TRAP binding to RNAs with between 2 and 11 GAGs, UAGs, AAGs, or CAGs showed that the stability of a TRAP-RNA complex is not directly proportional to the number of repeats in the RNA. These studies also showed that the effect of the identity of the residue in the first position of the triplet, with regard to binding to TRAP, is dependent on the number of repeats in the RNA. Together these data support a model in which TRAP binds to RNA by first forming an initial complex with a small subset of the repeats followed by a cooperative interaction with the remaining triplets.     

Determination of protein-DNA binding constants and specificities from statistical analyses of single molecules: MutS-DNA interactions

Atomic force microscopy (AFM) is a powerful technique for examining the conformations of protein–DNA complexes and determining the stoichiometries and affinities of protein–protein complexes. We extend the capabilities of AFM to the determination of protein–DNA binding constants and specificities. The distribution of positions of the protein on the DNA fragments provides a direct measure of specificity and requires no knowledge of the absolute binding constants. The fractional occupancies of the protein at a given position in conjunction with the protein and DNA concentrations permit the determination of the absolute binding constants. We present the theoretical basis for this analysis and demonstrate its utility by characterizing the interaction of MutS with DNA fragments containing either no mismatch or a single mismatch. We show that MutS has significantly higher specificities for mismatches than was previously suggested from bulk studies and that the apparent low specificities are the result of high affinity binding to DNA ends. These results resolve the puzzle of the apparent low binding specificity of MutS with the expected high repair specificities. In conclusion, from a single set of AFM experiments, it is possible to determine the binding affinity, specificity and stoichiometry, as well as the conformational properties of the protein–DNA complexes.     

Energetics of galactose- and glucose-aromatic amino acid interactions: implications for binding in galactose-specific proteins

An aromatic amino acid is present in the binding site of a number of sugar binding proteins. The interaction of the saccharide with the aromatic residue is determined by their relative position as well as orientation. The position-orientation of the saccharide relative to the aromatic residue was found to vary in different sugar-binding proteins. In the present study, interaction energies of the complexes of galactose (Gal) and of glucose (Glc) with aromatic residue analogs have been calculated by ab initio density functional (U-B3LYP/ 6-31G**) theory. The position-orientations of the saccharide with respect to the aromatic residue observed in various Gal-, Glc-, and mannose-protein complexes were chosen for the interaction energy calculations. The results of these calculations show that galactose can interact with the aromatic residue with similar interaction energies in a number of position-orientations. The interaction energy of Gal-aromatic residue analog complex in position-orientations observed for the bound saccharide in Glc/Man-protein complexes is comparable to the Glc-aromatic residue analog complex in the same position-orientation. In contrast, there is a large variation in interaction energies of complexes of Glc- and of Gal- with the aromatic residue analog in position-orientations observed in Gal-protein complexes. Furthermore, the conformation wherein the O6 atom is away from the aromatic residue is preferred for the exocyclic -CH2OH group in Gal-aromatic residue analog complexes. The implications of these results for saccharide binding in Gal-specific proteins and the possible role of the aromatic amino acid to ensure proper positioning and orientation of galactose in the binding site have been discussed.     

Cofilin binding to muscle and non-muscle actin filaments: isoform-dependent cooperative interactions

I have monitored equilibrium binding of human cofilin to rabbit skeletal muscle (alpha) and human non-muscle (85% beta, 15% gamma) actin filaments from the quenching of pyrene actin fluorescence. Filament binding is cooperative and stoichiometric (i.e. one cofilin molecule per actin subunit) for both actin isoforms. The Hill coefficient for binding to betagamma-actin filaments (n(H)=3.5) is greater than for muscle actin (n(H)=2.3). Analysis of equilibrium binding using a nearest-neighbor cooperativity model indicates that the intrinsic affinities for binding to an isolated site are comparable (10-14 microM) for both filament isoforms but the cooperative free energy is greater for binding betagamma-actin filaments. The predicted cofilin cluster sizes and filament binding densities are small at concentrations of cofilin where efficient filament severing is observed, indicating that a few bound cofilin molecules are sufficient to destabilize the filament lattice and promote fragmentation. The analysis used in this study provides a framework for evaluating proton and ion linkage and effects of regulatory proteins on cofilin binding and severing of actin filaments.     

Energetics of charge-charge interactions between residues adjacent in sequence

Statistical analysis of the residue separation between a pair of ionizable side chains within 4 ? of each other was performed on a set of 1560 non-homologous PDB structures. We found that the frequency of pairs of like charges (i.e., pairs consisting of acidic residues Asp and Glu or pairs consisting of basic residues Arg and Lys) is two orders of magnitude lower than the pairs of oppositely charged residues (salt-bridges). We also found that for pairs of like charges the distribution is skewed dramatically towards short residue separation (<3). On the basis of these observations, we hypothesize that at short residue separation the repulsion between charges does not contribute much to the protein stability and the effects are largely dominated by the long range charge-charge interactions with other ionizable groups in the protein molecule. To test this hypothesis, we incorporated various pairs of charged residues at position 63 and 64 of ubiquitin and compared the stabilities of these variants. We also performed calculations of the expected changes in the charge-charge interactions. A very good correlation between experimental changes in the stability of ubiquitin variants, and changes in the energy of charge-charge interactions provides support for the hypothesis that a pair of ionizable residues next to each other in sequence modulates protein stability via long range charge-charge interactions with the rest of the protein.