Roles of electrostatic interaction in proteins

Double mutant cycle analysis of aspartate 69, 97, and 103 to asparagine mutants in the m2 muscarinic acetylcholine receptor

Double mutant cycles provide a method for analyzing the effects of a mutation at a defined position in the protein structure on the properties of an amino acid at a second site. This approach was used to map potential interactions between aspartates 69, 97, and 103 in the m2 muscarinic acetylcholine receptor transmembrane helices 2 and 3. Receptors containing single and double aspartate to asparagine mutants were expressed in Chinese hamster ovary cells and their effects on ligand binding, signal transduction, and thermal stability determined. Analysis of the double mutant cycles showed that the mutations had approximately additive effects on ligand binding, signal transduction, and thermal stability. Ligand binding and thermal inactivation results support the conclusion that aspartate-103 is the ligand amine counterion. Effector coupling properties of the mutant receptors showed that aspartate-103 was also required for signal transduction activity. The mutation of aspartate-69 to asparagine completely eliminated signal transduction by the agonists acetylcholine, carbachol, and pilocarpine but not oxotremorine M, which caused reduced but significant inhibition of adenylyl cyclase and stimulation of phospholipase C. In contrast, adenylyl cyclase stimulation by the asparagine-69 mutant was elicited only by acetylcholine and carbachol but not by oxotremorine M. The variation in agonist-dependent effector coupling properties provides evidence that the asparagine-69 mutant can exist in activated receptor states that are different from the wild-type m2 muscarinic receptor.

Surface salt bridges stabilize the GCN4 leucine zipper

We present a study of the role of salt bridges in stabilizing a simplified tertiary structural motif, the coiled-coil. Changes in GCN4 sequence have been engineered that introduce trial patterns of single and multiple salt bridges at solvent exposed sites. At the same sites, a set of alanine mutants was generated to provide a reference for thermodynamic analysis of the salt bridges. Introduction of three alanines stabilizes the dimer by 1.1 kcal/mol relative to the wild-type. An arrangement corresponding to a complex type of salt bridge involving three groups stabilizes the dimer by 1.7 kcal/ mol, an apparent elevation of the melting temperature relative to wild type of about 22 degrees C. While identifying local from nonlocal contributions to protein stability is difficult, stabilizing interactions can be identified by use of cycles. Introduction of alanines for side chains of lower helix propensity and complex salt bridges both stabilize the coiled-coil, so that combining the two should yield melting temperatures substantially higher than the starting species, approaching those of thermophilic sequences.

Electrostatic interactions drive scaffolding/coat protein binding and procapsid maturation in bacteriophage P22

The first step in assembly of the bacteriophage P22 is the formation of a T=7 icosahedral "procapsid," the major components of which are the coat protein and an inner core composed of the scaffolding protein. Although not present in the mature virion, the scaffolding protein is required for procapsid assembly. Eleven amino-acid residues at the extreme carboxyl terminus of the scaffolding protein are required for binding to the coat protein, and upon deletion of these residues, approximately 20 additional residues become disordered. Sequence analysis and NMR data suggest that the 30 residues at the carboxyl terminus form a helix-loop-helix motif which is stabilized by interhelical hydrophobic interactions. This "coat protein recognition domain" presents an unusually high number of positively charged residues on one face, suggesting that electrostatic interactions between this domain and the coat protein may contribute to recognition and binding. We report here that high ionic strength (1 M NaCl) completely inhibited procapsid assembly in vitro. When scaffolding protein was added to empty procapsid "shells" of coat protein, 1 M NaCl partially inhibited the binding of scaffolding protein to the shells. This suggests that the positively charged coat protein recognition domain at the carboxyl terminus of the scaffolding protein binds to a negatively charged region on the coat protein. During DNA packaging, the scaffolding protein exits the procapsid; scaffolding protein exit is followed by the expansion of the procapsid into a mature capsid. Procapsid shells can be induced to undergo a similar expansion reaction in vitro by heating (45-70 degreesC); this process was also inhibited by 1 M NaCl. These results are consistent with a model in which negatively charged scaffold protein-binding domains in the coat proteins move apart during procapsid expansion; this relief of electrostatic repulsion could provide a driving force for expansion and subsequent maturation. High-salt concentrations would screen this repulsion, while packaging of DNA (a polyanion) in vivo may increase the instability of the procapsid enough to trigger its expansion.

In vitro heat effect on heterooligomeric subunit assembly of thermostable indolepyruvate ferredoxin oxidoreductase

Indolepyruvate ferredoxin oxidoreductase (IOR) from hyperthermophilic archaeon Pyrococcus kodakaraensis KOD1 catalyzes the oxidative decarboxylation of arylpyruvates by forming a heterooligomeric complex (alpha2beta2). The genes iorA and iorB which encode respective alpha and beta subunits, were coexpressed heterologously in Escherichia coli cells under anaerobic conditions. IOR activity was detected from the cell extract containing both subunits and its activity was enhanced by in vitro heat treatment prior to the assay. The iorA and iorB were expressed individually and each subunit was examined for enzymatic activity with and without heat treatment. IOR activity was detected neither from the extract of alpha subunit nor beta subunit. The alpha and beta subunits were mixed and then IOR activity was examined. Weak IOR activity was detected without heat treatment, however, upon heat treatment its activity was enhanced. The mixture of individually heat treated alpha and beta subunits did not possess any IOR activity even though the mixed sample was heat treated again. IOR alpha and beta subunits were individually purified to homogeneity, mixed with or without heat treatment and subunit assembly was examined by determining molecular mass. Upon heat treatment, inactive alpha and beta were converted to an active high molecular weight complex (195 kDa) which corresponds to the alpha2beta2 structure. However, the active complex was not formed without heat treatment, suggesting that high temperature environments are important for the heterooligomerization of IOR subunits.

Ionic strength-dependent physicochemical factors in cytochrome c3 regulating the electron transfer rate

The effect of ionic strength on the macroscopic and microscopic redox potentials and the heme environment of cytochrome c3 from Desulfovibrio vulgaris Miyazaki F have been investigated by NMR and electrochemical methods. The redox potentials of this tetraheme protein are found to be ionic strength-dependent. Especially, the microscopic redox potentials of hemes 2 and 3 at the fourth reduction step increase significantly with increasing ionic strength, which is in contraction to the theoretical expectation. The coordinated imidazole proton signals are unaffected by ionic strength. However, the methyl and propionate proton signals of hemes 1 and 4 showed significant ionic strength dependencies that are distinct from those for hemes 2 and 3. This heme classification is the same as that found in the ionic strength dependencies of the microscopic redox potentials at the fourth reduction step. Furthermore, the effect of ionic strength on the electrostatic potentials at the heme irons has been examined on the theoretical basis. The electrostatic potential at heme 4 changes up to 1 M ionic strength, which was not expected from the observations reported on cytochromes so far. These results are discussed in connection with the reported anomalous ionic strength dependency of the reduction rate of cytochrome c3.

Intrinsic protein electric fields: basic non-covalent interactions and relationship to protein-induced Stark effects

Knowledge of the interactions involving charged, polar and polarizable groups in proteins is fundamental, not only because they are important determinants for gaining insight into biophysical molecular recognition and assembly processes, but also for understanding how the matrix of a protein can be viewed as an electric field capable of inducing Stark perturbations on the spectral properties of biological optical centers. This review describes the essential features of noncovalent interactions in protein systems and discusses the concept of the dielectric constant of a protein in the context of different microscopic and macroscopic modeling approaches. It also provides an account of a specific type of high resolution vibrational and optical Stark spectroscopy attempting to correlate the observed spectral properties of biological optical centers to the intrinsic protein fields induced by the matrix in which they reside.

Hydration and conformational equilibria of simple hydrophobic and amphiphilic solutes

We consider whether the continuum model of hydration optimized to reproduce vacuum-to-water transfer free energies simultaneously describes the hydration free energy contributions to conformational equilibria of the same solutes in water. To this end, transfer and conformational free energies of idealized hydrophobic and amphiphilic solutes in water are calculated from explicit water simulations and compared to continuum model predictions. As benchmark hydrophobic solutes, we examine the hydration of linear alkanes from methane through hexane. Amphiphilic solutes were created by adding a charge of +/-1e to a terminal methyl group of butane. We find that phenomenological continuum parameters fit to transfer free energies are significantly different from those fit to conformational free energies of our model solutes. This difference is attributed to continuum model parameters that depend on solute conformation in water, and leads to effective values for the free energy/surface area coefficient and Born radii that best describe conformational equilibrium. In light of these results, we believe that continuum models of hydration optimized to fit transfer free energies do not accurately capture the balance between hydrophobic and electrostatic contributions that determines the solute conformational state in aqueous solution.

Ion pairs involved in maintaining a thermostable structure of glutamate dehydrogenase from a hyperthermophilic archaeon

Intersubunit ion pairs are considered to be involved for maintaining a stable structure of the glutamate dehydrogenase (GDH) from hyperthermophiles. In order to demonstrate an effect of intersubunit ion pairs on the structural stability, two kinds of mutation (T138E, Thr at position 138 was replaced by Glu; E158Q, Glu at position 158 was replaced by Gln) which add and remove ion pairs, respectively, were introduced into Pk-gdhA gene encoding GDH from Pyrococcus kodakaraensis KOD1. Addition of one ion pair (Pk-GDHA-T138E) increased the optimum temperature and thermostability. In contrast, Pk-GDH-E158Q showed lower optimum temperature and less thermostability than wild type GDH. Structure analysis of GDHs was performed by circular dichroism (CD) and indicated that all recombinant enzymes (Pk-GDH, Pk-GDH-T138E, Pk-GDH-E158Q) possess different structures from that of natural GDH. Upon heat treatment (60 degrees C, 2 h), the structures of Pk-GDH and Pk-GDH-T138E were converted to another form close to the natural structure. However, no structural conversion by heat treatment was observed in Pk-GDH-E158Q. These results indicate that intersubunit ion pairs play an important role in forming thermostable structure of Pk-GDH.

Charge and solvation effects in anion recognition centers: an inquiry exploiting reactive arginines

Following a long-standing suggestion of Riordan et al. [Riordan, J. F., McElvany, K. D., and Borders, C. L., Jr. (1977) Science 195, 884-885], we sought to exploit chemically activated arginines as probes to characterize the microenvironmental effects in enzymes that mediate the recognition of anionic substrates. A micellar simulation study establishes that octylguanidine (OGn) becomes chemically activated upon incorporation into both cetyltrimethylammonium bromide (CTAB) and Triton X-100 micelles and that the activations correlate with the pKa diminutions induced in its guanidinium group by the effects of electrostatic or nonelectrostatic nature as reflected in the results of pH and salt titration experiments. Next, a protein modification study establishes that the modifiable arginines in a number of enzymes also have diminished pKa's, again due to effects of electrostatic or nonelectrostatic nature as reflected in the results of pH and salt titration experiments. Warwicker's finite difference Poisson--Boltzmann algorithm [Warwicker, J. (1992) J. Mol. Biol. 223, 247-257] is applied to several of the enzymes with available crystal structure coordinates, and indeed, their chemically activated arginines are found to be in an electrostatic microenvironment that can diminish their pKa's, with the magnitudes of these diminutions matching closely the diminutions measured experimentally. Finally, the chemically activated arginines are examined with respect to their atomic atmosphere and are thus found to occur in a local microenvironment that would facilitate their roles as anion anchors. Thus, electrostatic and solvation effects are found to be critical determinants of the arginine role as an anion anchor.

FTIR study of the thermal denaturation of alpha-actinin in its lipid-free and dioleoylphosphatidylglycerol-bound states and the central and N-terminal domains of alpha-actinin in D2O

Fourier transform infrared (FTIR) spectroscopy has been carried out to investigate the thermal denaturation of alpha-actinin and its complexes with dioleoylphosphatidylglycerol (DOPG) vesicles. The amide I regions in the deconvolved spectra of alpha-actinin in the lipid-free and DOPG-bound states are both consistent with predominantly alpha-helical secondary structure below the denaturation temperatures. Studies of the temperature dependence of the spectra revealed that for alpha-actinin alone the secondary structure was unaltered up to 40 degrees C. But, in the presence of DOPG vesicles, the thermal stability of the secondary structure of alpha-actinin increased to 55 degrees C. The thermal denaturation mechanisms of the lipid-free and DOPG-bound states of alpha-actinin also vary. The secondary structure of the lipid-free alpha-actinin changed to be predominantly unordered upon heating to 65 degrees C and above. Whereas, the original alpha-helical structure in the DOPG-bound alpha-actinin retained even at 70 degrees C, the highest temperature we examined. Analysis of the reduction in amide II intensities, which is due to peptide H-D exchange upon heating alpha-actinin in D2O, showed that partially unfolded states with increased solvent accessibility but substantial secondary structures could be observed from 35 to 40 degrees C only if DOPG vesicles were present. A so-called "protamine precipitation" method has been developed to purify the N-terminal domain of alpha-actinin by use of the fact that the central domain of alpha-actinin is negatively charged but the N-terminal domain is positively charged. Thermal denaturation of the central and N-terminal domains of alpha-actinin were then investigated with FTIR. The secondary structure of the N-terminal domain of alpha-actinin was found to be thermally sensitive below 35 degrees C, which is characterized as the increase of the alpha-helical structure at the expense of the random coil upon heating the N-terminal domain from 4 to 35 degrees C. The membrane-binding ability of the N-terminal domain of alpha-actinin was proposed in terms of the analysis of the local electrostatic properties of alpha-actinin and the assignment of the amide II bands in the FTIR spctra of alpha-actinin.

Direct visualization of surface charge in aqueous solution

Tapping mode atomic force microscopy operated in the phase shift mode is used to directly visualize the surface charge on biological macromolecules under solution. A simple theory is presented which is qualitatively consistent with experimental observations, although a more complicated theoretical analysis is required for a quantitative comparison.

Electrostatic stabilization in methionine aminopeptidase from hyperthermophile Pyrococcus furiosus

The thermostability of methionine aminopeptidase from a hyperthermophile P. furiosus (PfMAP) was extremely high: the denaturation temperature was 106.2 degreesC at pH 10.2. To explore the contribution of electrostatic interaction to the superior thermostability of PfMAP, the thermostability of PfMAP was examined by differential scanning calorimetry (DSC) in various salt concentrations in the acidic region far from the isoelectric point of PfMAP. (1) In 20 mM glycine buffer, the DSC curve of PfMAP exhibited a single peak. Transition temperatures (Tm) were lowered with decreasing pH from 4 to 3. The heat denaturation of PfMAP was not reversible. (2) Denaturation enthalpy (DeltaH) measured at different pHs linearly correlated with Tm up to 102 degreesC, suggesting that the denaturation heat capacity (DeltaCp) for PfMAP is constant up to 100 degreesC. DeltaCp was estimated to be 0.82 J K-1 g-1. (3) In the presence of 10-100 mM KCl at pH 3.2, two peaks appeared on the DSC curves. The first peak shifted to lower temperatures with increasing concentration of KCl and, oppositely, the second one to higher temperatures. It was found that the first and second peaks originated from the heat denaturation of the native form of PfMAP and the melting of the non-native associated form having molten globule-like structure, respectively, judged from the CD spectra and ultracentrifugation analyses. This indicates the following: first, the attractive electrostatic interaction is an important factor in stabilizing the native form of PfMAP; second, the presence of KCl stimulates the formation of the molten globule-like state of PfMAP and stabilizes it. (4) In a comparison of the sequence and crystal structure of PfMAP, which has been recently determined (1xgs.pdb), with those of MAP from Escherichia coli (EcMAP), it was predicted that the extra four short-range ion pairs less than 3 A involved in PfMAP are crucial candidates as determinants for the superior thermostability of PfMAP.

The effect of protein relaxation on charge-charge interactions and dielectric constants of proteins

The effect of the reorganization of the protein polar groups on charge-charge interaction and the corresponding effective dielectric constant (epsilon(eff)) is examined by the semimicroscopic version of the Protein Dipole Langevin Dipoles (PDLD/S) method within the framework of the Linear Response Approximation (LRA). This is done by evaluating the interactions between ionized residues in the reaction center of Rhodobacter sphaeroides, while taking into account the protein reorganization energy. It is found that an explicit consideration of the protein relaxation leads to a significant increase in epsilon(eff) and that semimicroscopic models that do not take this relaxation into account force one to use a large value for the so-called "protein dielectric constant," epsilon(p), of the Poisson-Boltzmann model or for the corresponding epsilon(in) in the PDLD/S model. An additional increase in epsilon(eff) is expected from the reorganization of ionized residues and from changes in the degree of water penetration. This finding provides further support for the idea that epsilon(in) (or epsilon(p)) represents contributions that are not considered explicitly. The present study also provides a systematic illustration of the nature of epsilon(eff), supporting our previously reported view that charge-charge interactions correspond to a large value of this "dielectric constant," even in protein interiors. It is also pointed out that epsilon(eff) for the interaction between ionizable groups in proteins is very different from the effective dielectric constant, epsilon'(eff), that determines the free energy of ion pairs in proteins (epsilon'(eff) reflects the effect of preoriented protein dipoles). Finally, the problems associated with the search for a general epsilon(in) are discussed. It is clarified that the epsilon(in) that reproduces the effect of protein relaxation on charge-charge interaction is not equal to the epsilon(in) that reproduces the corresponding effect upon formation of individual charges. This reflects fundamental inconsistencies in attempts to cast microscopic concepts in a macroscopic model. Thus one should either use a large epsilon(in) for charge-charge interactions and a small epsilon(in) for charge-dipole interactions or consider the protein relaxation microscopically.

Electrostatic effects in macromolecules: fundamental concepts and practical modeling

The past few years have seen an exponential growth in the calculations of electrostatic energies of macromolecules and an increased recognition of the crucial role of electrostatic effects. This review considers the current state of the field. Focus is placed on calculations of pKas, redox potentials and binding energies in macromolecules and clarification of the fact that the value of the dielectric 'constant' of a protein depends on its definition and that small dielectric constants should not be used in describing charge-charge interactions by current continuum models.

Engineering of a stable mutant malic enzyme by introducing an extra ion-pair to the protein

A double mutant (R9E/M17K) of pigeon liver malic enzyme with glutamate and lysine replaced for arginine and methionine at positions 9 and 17, respectively, was found to be much more stable in urea and thermal denaturation, but was enzymatically less active than the wild-type enzyme (WT). Unfolding of the enzyme by urea produced a large red shifting of the protein fluorescence maximum from 320 to 360 nm, which was completely reversible upon dilution. Analysis of the denaturation curves monitored by enzyme activity lost suggested that a putative intermediate was involved in the denaturation process. The half unfolding urea concentration, measured by fluorescence spectral changes, increased from 2.24 M for WT to 3.13 M for R9E/M17K. The melting temperature increased by approximately 10 degrees C for R9E/M17K compared with that for WT. Kinetic analysis of the thermal inactivation at 58 degrees C also conformed to a three-state model with the rate constant for the intermediate state of R9E/M17K (k2 = 0.03 min(-1)) being much smaller than the WT value (k2 = 2.39 min(-1)). Results obtained from single mutants indicated that the decreasing enzyme activity of R9E/M17K was exclusively due to R9 mutation, which increased the K(mMn) and K(mMal) by at least one order of magnitude compared with WT. Consequently, a decrease occurred in the specificity constant [k(cat)/(K(mMm)K(mNADP)K(mMal))] for the R9 mutants at least four orders of magnitude smaller than the WT. M17K has similar properties to the WT, while R9E is more labile than the WT enzyme. The above results indicate that the extra stability gained by the double mutant possibly occurs through the introduction of an extra ion-pair between E9 and K17, which freezes the double mutant in the putative intermediate state. Examination of the N-terminal amino acid sequence of pigeon liver malic enzyme reveals that position 15 is also a lysine residue. Since the R9E mutant, which has an extra Glu9-Lys15 ion-pair, is less stable than the WT, we conclude that the contribution to malic enzyme stability is specific for the Glu9-Lys17 ion-pair.

pH-dependent gating in the Streptomyces lividans K+ channel

Because of its size, high levels of expression, and unusual detergent stability, the small K+ channel from Streptomyces lividans (SKC1) is considered to be an ideal candidate for detailed structural analysis. In this paper, we have used planar lipid bilayers and radiotracer uptake experiments to study purified and reconstituted SKC1, in an attempt to develop a bulk assay for its functional characterization. In channels reconstituted into liposomes with external pH 3.5 and intravesicular pH 7.5, a time-dependent SKC1-catalyzed 86Rb+ uptake was observed. This cationic influx was blocked by Ba2+ ions with a Ki (external) of 0.4 mM and was shown to have the following selectivity sequence: K+ > Rb+ > NH4+ >> Na+ > Li+. In experiments with external pH 7.5 or in liposomes containing no channels, no 86Rb+ uptake was detected. When SKC1 was incorporated into planar lipid bilayers, we failed to observe significant single-channel activity at neutral pH but detected frequent multiple-channel openings a pH < 5.0. These results indicate that under these experimental conditions SKC1 behaves as a pH-gated K+ channel in which protonation of one or more residues promotes channel opening. At acidic pH and symmetrical 200 mM KCl solutions, SKC1 showed numerous brief openings with a main single-channel conductance of 135 pS and a subconductance state of 70 pS. Channel open probability showed a slight voltage dependence, with higher activities observed at negative potentials, a fact which may suggest that the protonation site lies within the transmembrane electrical field. Attempts to determine the pKa of channel activation were obscured by intrinsic limitations of the 86Rb+ flux assay. However, it appears to be lower than pH 4.0. Limited proteolysis experiments demonstrated that SKC1 reconstitutes vectorially, almost exclusively in the right-side-out configuration, indicating that the protonation site responsible for channel opening is located at the extracellular face of the channel. These results point toward a potentially novel gating mechanism for SKC1 and open the possibility of using transmembrane-driven radiotracer influx experiments as a reliable bulk functional assay for reconstituted SKC1.

Thermodynamic and kinetic basis of interfacial activation: resolution of binding and allosteric effects on pancreatic phospholipase A2 at zwitterionic interfaces

A general kinetic model for catalysis by interfacial enzymes is developed. It couples the Michaelis-Menten catalytic turnover cycle at the interface with that in the aqueous phase through the distribution equilibria between the interface and the surrounding aqueous phase. Analysis under two limiting conditions fully describes the steady-state kinetics of hydrolysis and resolves the allosteric effects from apparent modes of interfacial activation in terms of the primary rate and equilibrium parameters for pig pancreatic phospholipase A2 (PLA2). One limit is observed in dispersions of anionic phospholipid vesicles, in which intervesicle exchange of enzyme, substrate, and hydrolysis products is absent and reaction occurs only on vesicles containing enzyme. A complete analysis at this highly processive limit, called kinetics in the scooting mode, has been published [Berg et al. (1991) Biochemistry 30, 7283]. Here is reported the analysis in the other limit, PLA2-catalyzed hydrolysis of zwitterionic micelles of short-chain phosphatidylcholines, at which substrate and products are in rapid exchange. Hydrolysis occurs either in bulk aqueous solution with phospholipid monomers or at the micellar interface. Above the critical micelle concentration (cmc), the hydrolysis rate shows a hyperbolic dependence on the bulk substrate concentration present as micelles. This dependence, characterized by the fitting parameters KMapp and VMapp, is analyzed in terms of the primary rate and equilibrium constants. The kinetic analysis is based on the assumption that the microscopic steady-state condition is satisfied because substrate replenishment in the micro-environment of the enzyme is fast relative to the catalytic turnover time. Added NaCl and anionic interface increase the hydrolysis rate in zwitterionic micelles dramatically. The overall interfacial rate enhancement is attributed to three factors: (a) promotion of PLA2 binding by net anionic charge of the interface, (b) enhancement of substrate affinity of PLA2 at the interface (Ks* allostery), and (c) stimulation of the rate-limiting chemical step (kcat* allostery).

Protein stability; optimization of electrostatic contributions by partially neutralizing surface ionic charges

'Partial Charge-Neutralization Method' is developed to study influence of the relative amount of positive and negative charges in proteins on their structural stability. A given number of either positively or negatively charged groups are neutralized in all of their possible combinations to generate a whole set of distinct species. The Coulomb energy of each species is calculated by numerically solving the Poisson-Boltzmann equation for aqueous solutions. Partial neutralization of lysine residues of tuna cytochrome c in aqueous solution at neutral pH with the Debye-Hückel screening parameter kappa = 1 nm-1 reproduces qualitatively well the destabilization of acetylated cytochrome c observed in physicochemical measurements at pH 7. The stabilization of its molten globule state at pH 2 is also studied with the present method. It is shown that the electrostatic contribution to the structural stability of natural proteins can be optimized by changing the difference in number of their positive and negative charges.

Non-random ionic-charge distribution responsible for the structural stability and molecular recognition of proteins

The 'ionic-charge shuffling method' is presented to generate a complete set of electrostatic mutants for a natural protein where ionic charges on the molecular surface of the template protein are exhaustively interchanged with each other. Total Coulomb interaction energies are evaluated for all of the mutants by numerically solving the finite difference Poisson-Boltzmann equation and their distribution in the ensemble is obtained. This method has been applied to five natural proteins to reveal that they have a significantly lower Coulomb energy than the average over the ensemble of their mutants. It is also shown that these natural proteins have a significantly larger and smaller number of pairs of attractive and repulsive ionic groups, respectively, than those expected for their randomly shuffled ensemble: They have been 'designed' through molecular evolution so that a pair of ionic charges with opposite signs may have a higher tendency to be located close to each other, while a pair with the same sign are away from each other.

Backbone makes a significant contribution to the electrostatics of alpha/beta-barrel proteins

The electrostatic properties of seven alpha/beta-barrel enzymes selected from different evolutionary families were studied: triose phosphate isomerase, fructose-1,6-bisphosphate aldolase, pyruvate kinase, mandelate racemase, trimethylamine dehydrogenase, glycolate oxidase, and narbonin, a protein without any known enzymatic activity. The backbone of the alpha/beta-barrel has a distinct electrostatic field pattern, which is dipolar along the barrel axis. When the side chains are included in the calculations the general effect is to modulate the electrostatic pattern so that the electrostatic field is generally enhanced and is focused into a specific area near the active site. We use the electrostatic flux through a square surface near the active site to gauge the functionally relevant magnitude of the electrostatic field. The calculations reveal that in six out of the seven cases the backbone itself contributes greater than 45% of the total flux. The substantial electrostatic contribution of the backbone correlates with the known preference of alpha/beta-barrel enzymes for negatively charged substrates.

The crystal structure of citrate synthase from the hyperthermophilic archaeon pyrococcus furiosus at 1.9 A resolution,

The crystal structure of the closed form of citrate synthase, with citrate and CoA bound, from the hyperthermophilic Archaeon Pyrococcus furiosus has been determined to 1.9 A. This has allowed direct structural comparisons between the same enzyme from organisms growing optimally at 37 degrees C (pig), 55 degrees C (Thermoplasma acidophilum) and now 100 degrees C (Pyrococcus furiosus). The three enzymes are homodimers and share a similar overall fold, with the dimer interface comprising primarily an eight alpha-helical sandwich of four antiparallel pairs of helices. The active sites show similar modes of substrate binding; moreover, the structural equivalence of the amino acid residues implicated in catalysis implies that the mechanism proceeds via the same acid-base catalytic process. Given the overall structural and mechanistic similarities, it has been possible to make detailed structural comparisons between the three citrate synthases, and a number of differences can be identified in passing from the mesophilic to thermophilic to hyperthermophilic citrate synthases. The most significant of these are an increased compactness of the enzyme, a more intimate association of the subunits, an increase in intersubunit ion pairs, and a reduction in thermolabile residues. Compactness is achieved by the shortening of a number of loops, an increase in the number of atoms buried from solvent, an optimized packing of side chains in the interior, and an absence of cavities. The intimate subunit association in the dimeric P. furiosus enzyme is achieved by greater complementarity of the monomers and by the C-terminal region of each monomer folding over the surface of the other monomer, in contrast to the pig enzyme where the C-terminus has a very different fold. The increased number of intersubunit ion pairs is accompanied by an increase in the number involved in networks. Interestingly, all loop regions in the P. furiosus enzyme either are shorter or contain additional ion pairs compared with the pig enzyme. The possible relevance of these structural features to enzyme hyperthermostability is discussed.

Design and characterization of the anion-sensitive coiled-coil peptide

As a model for analyzing the role of charge repulsion in proteins and its shielding by the solvent, we designed a peptide of 27 amino acid residues that formed a homodimeric coiled-coil. The interface between the coils consisted of hydrophobic Leu and Val residues, and 10 Lys residues per monomer were incorporated into the positions exposed to solvent. During the preparation of a disulfide-linked dimer in which the two peptides were linked in parallel by the two disulfide bonds located at the N and C terminals, a cyclic monomer with an intramolecular disulfide bond was also obtained. On the basis of CD and 1H-NMR, the conformational stabilities of these isomers and several reference peptides were examined. Whereas all these peptides were unfolded in the absence of salt at pH 4.7 and 20 degrees C, the addition of NaClO4 cooperatively stabilized the alpha-helical conformation. The crosslinking of the peptides by disulfide bonds significantly decreased the midpoint salt concentration of the transition. The 1H-NMR spectra in the presence of NaClO4 suggested that, whereas the disulfide-bonded dimer assumed a native-like conformation, the cyclic monomer assumed a molten globule-like conformation with disordered side chains. However, the cyclic monomer exhibited cooperative transitions against temperature and Gdn-HCl that were only slightly less cooperative than those of the disulfide-bonded parallel dimer. These results indicate that the charge repulsion critically destabilizes the native-like state as well as the molten globule-like state, and that the solvent-dependent charge repulsion may be useful for controlling the conformation of designed peptides.

The effect of ionic strength and specific anions on substrate binding and hydrolytic activities of Na,K-ATPase

The physiological ligands for Na,K-ATPase (the Na,K-pump) are ions, and electrostatic forces, that could be revealed by their ionic strength dependence, are therefore expected to be important for their reaction with the enzyme. We found that the affinities for ADP3-, eosine2-, p-nitrophenylphosphate, and V(max) for Na,K-ATPase and K+-activated p-nitrophenylphosphatase activity, were all decreased by increasing salt concentration and by specific anions. Equilibrium binding of ADP was measured at 0-0.5 M of NaCl, Na2SO4, and NaNO3 and in 0.1 M Na-acetate, NaSCN, and NaClO4. The apparent affinity for ADP decreased up to 30 times. At equal ionic strength, I, the ranking of the salt effect was NaCl approximately Na2SO4 approximately Na-acetate < NaNO3 < NaSCN < NaCl04. We treated the influence of NaCl and Na2SO4 on K(diss) for E x ADP as a "pure" ionic strength effect. It is quantitatively simulated by a model where the binding site and ADP are point charges, and where their activity coefficients are related to I by the limiting law of Debye and Hückel. The estimated net charge at the binding site of the enzyme was about +1. Eosin binding followed the same model. The NO3- effect was compatible with competitive binding of NO3- and ADP in addition to the general I-effect. K(diss) for E x NO3 was approximately 32 mM. Analysis of V(max)/K(m) for Na,K-ATPase and K+-p-nitrophenylphosphatase activity shows that electrostatic forces are important for the binding of p-nitrophenylphosphate but not for the catalytic effect of ATP on the low affinity site. The net charge at the p-nitrophenylphosphate-binding site was also about +1. The results reported here indicate that the reversible interactions between ions and Na,K-ATPase can be grouped according to either simple Debye-Hückel behavior or to specific anion or cation interactions with the enzyme.

Bioactivities and secondary structures of constrained analogues of human parathyroid hormone: cyclic lactams of the receptor binding region

In a search for analogues of human parathyroid hormone (hPTH) with improved activities and bioavailabilities, we have prepared the following three lactam analogues of hPTH-(1-31)-NH2 (1) or [Leu27]hPTH-(1-31)-NH2 (2): [Leu27]cyclo(Glu22-Lys26)-hPTH-(1-31)-NH2 (3), [Leu27]cyclo(Lys26-Asp30)-hPTH-(1-31)-NH2 (4), and cyclo(Lys27-Asp30)-hPTH-(1-31)-NH2 (5). Analogues 1, 2, and 5 had seven or eight residues of alpha-helix, as estimated from their circular dichroism (CD) spectra, in contrast to 12 residues in cyclic analogues 3 and 4. Thus, lactams 3 and 4 stabilized a helix previously shown to exist within residues 17-29. The adenylyl cyclase activity (EC50), measured in rat osteosarcoma 17/2 cells, of 5 (40.3 +/- 2.3 nM) was half that of its linear form 1 (19.9 +/- 3.9 nM). The linear Leu27 mutant 2 was twice as active (11.5 +/- 5.2) as analogue 1, and lactam analogue 3 was 6-fold more active (3.3 +/- 0.3 nM). Lactam analogue 4 had less activity (16.9 +/- 3.3 nM) than 2, its linear form. Peptides hPTH-(1-30)-NH2 (6), [Leu27]hPTH-(1-30)-NH2 (7), and [Leu27]cyclo(Glu22-Lys26)-hPTH-(1-30)-NH2 (8) all had AC-stimulating activities similar to that of 1. When injected intravenously, with a dose of 0.8 nmol/100 g of analogue in acid saline, hypotensive effects paralleled their adenylyl cyclase activities. They behaved quite differently when applied subcutaneously. Analogues 1, 5, and 6, the weakest, showed about half the drop in blood pressure observed with 3 and 4, the most active. In contrast, the time required to reach a maximum drop in blood pressure of 4-8, after subcutaneous administration, was 2-4 times that of the other analogues. Thus, the bioavailabilities of the lactam analogues, unlike their adenylyl cyclase-stimulating activities, were highly dependent on the presence or conformation of Val31.

Electrostatic interactions between transmembrane segments mediate folding of Shaker K+ channel subunits

In voltage-dependent Shaker K+ channels, charged residues E293 in transmembrane segment S2 and R365, R368, and R371 in S4 contribute significantly to the gating charge movement that accompanies activation. Using an intragenic suppression strategy, we have now probed for structural interaction between transmembrane segments S2, S3, and S4 in Shaker channels. Charge reversal mutations of E283 in S2 and K374 in S4 disrupt maturation of the protein. Maturation was specifically and efficiently rescued by second-site charge reversal mutations, indicating that electrostatic interactions exist between E283 in S2 and R368 and R371 in S4, and between K374 in S4 and E293 in S2 and D316 in S3. Rescued subunits were incorporated into functional channels, demonstrating that a native structure was restored. Our data indicate that K374 interacts with E293 and D316 within the same subunit. These electrostatic interactions mediate the proper folding of the protein and are likely to persist in the native structure. Our results raise the possibility that the S4 segment is tilted relative to S2 and S3 in the voltage-sensing domain of Shaker channels. Such an arrangement might provide solvent access to voltage-sensing residues, which we find to be highly tolerant of mutations.

A small engineered protein lacks structural uniqueness by increasing the side-chain conformational entropy

A small globular protein, the third repeat of the c-Myb DNA-binding domain, which is composed of 54 amino acid residues, was engineered so as to understand the structural uniqueness of native proteins. This small protein has three alpha-helices that form a helix-turn-helix structure, which is maintained by the hydrophobic core with three Ile residues. One of the mutant proteins, with two of the buried Ile (Ile-155 and Ile-181) substituted with Leu residues, showed multiple conformations, as monitored by heteronuclear magnetic resonance spectroscopy for 13C- and 15N-labeled proteins. The increase in the side-chain conformational entropy, caused by changing the Ile to a Leu residue on an alpha-helix, could engender the lack of structural uniqueness. In native proteins, the conformations of not only the beta-branched side chains, but also those of the neighboring bulky side chains, can be greatly restricted, depending upon the local backbone structure.

Electrostatic interactions in the active site of the N-terminal thioredoxin-like domain of protein disulfide isomerase

Proteins with the thioredoxin fold have widely differing stabilities of the disulfide bond that can be formed between the two cysteines at their active site sequence motif Cys1-Xaa2-Yaa3-Cys4. This is believed to be regulated not by varying the disulfide bond itself, but by modulating the stability of the dithiol form of the protein through interactions with the ionized form of the Cys1 thiol group. A consistent relationship between disulfide bond stability and Cys1 thiol pKa value is found here for DsbA, thioredoxin, and the N-terminal thioredoxin-like domain of protein disulfide isomerase (PDI a), which has a very low thiol pKa value of 4.5. This thiolate anion is stabilized by 5.7 kcal/mol in the dithiol form, giving rise to the corresponding instability of the disulfide bond and the oxidizing properties of PDI a. Electrostatic interactions in the active site of the PDI a-domain have been characterized in order to understand the physical basis of this stabilization. Linkage with the ionization of the imidazole group of His3 in the active site demonstrates that this charge-charge interaction contributes 1.1 kcal/mol. The remainder of the stabilization is believed to be due primarily to interactions with the partial positive charges at the N-terminus of an alpha-helix, which are exceedingly sensitive to charges of surrounding residues.