The deuterium isotope effect on the NMR signal of the low-barrier hydrogen bond in a transition-state analog complex of chymotrypsin

The participation of a low-barrier hydrogen bond (LBHB) in the mechanism of action of chymotrypsin introduces a new role for Asp 102 and His 57 in catalysis [C. S. Cassidy, J. Lin, and P. A. Frey (1997) Biochemistry 36, 4576-4584]. It is postulated that the LBHB increases the basicity of His 57-N(epsilon2) in the transition state, thereby facilitating the abstraction of a proton from Ser 195, and stabilizes the tetrahedral intermediate in the acylation step. Evidence for this mechanism includes the downfield chemical shift of the proton bridging His 57 and Asp 102 in transition-state analog complexes and the low deuterium fractionation factors for this proton in the same complexes. We present additional spectroscopic evidence supporting the assignment of an LBHB between His 57 and Asp 102. The tetrahedral addition complex between Ser 195 of chymotrypsin and N-acetyl-l-leucyl-l-phenylalanyl trifluoromethylketone is regarded as a close structural analog of a tetrahedral intermediate. The deuterium NMR signal for the downfield deuteron bridging His 57 and Asp 102 in D(2)O has now been observed as a broad band centered at 17.8 +/- 0.5 ppm. The proton NMR signal in H(2)O is centered at 18.9 +/- 0.05 ppm. The two signals are clearly separated corresponding to a deuterium isotope effect of Delta[delta(H) - delta(D)] = 1.1 +/- 0.5 ppm. Deuterium isotope effects in this range are characteristic of LBHBs, and this observation provides further support for the assignment of the proton bridging His 57 and Asp 102 in transition-state analog complexes as an LBHB.     

Correlation of low-barrier hydrogen bonding and oxyanion binding in transition state analogue complexes of chymotrypsin

The structures of the hemiketal adducts of Ser 195 in chymotrypsin with N-acetyl-L-leucyl-L-phenylalanyl trifluoromethyl ketone (AcLF-CF3) and N-acetyl-L-phenylalanyl trifluoromethyl ketone (AcF-CF3) were determined to 1.4-1.5 A by X-ray crystallography. The structures confirm those previously reported at 1.8-2.1 A [Brady, K., Wei, A., Ringe, D., and Abeles, R. H. (1990) Biochemistry 29, 7600-7607]. The 2.6 A spacings between Ndelta1 of His 57 and Odelta1 of Asp 102 are confirmed at 1.3 A resolution, consistent with the low-barrier hydrogen bonds (LBHBs) between His 57 and Asp 102 postulated on the basis of spectroscopy and deuterium isotope effects. The X-ray crystal structure of the hemiacetal adduct between Ser 195 of chymotrypsin and N-acetyl-L-leucyl-L-phenylalanal (AcLF-CHO) has also been determined at pH 7.0. The structure is similar to the AcLF-CF3 adduct, except for the presence of two epimeric adducts in the R- and S-configurations at the hemiacetal carbons. In the (R)-hemiacetal, oxygen is hydrogen bonded to His 57, not the oxyanion site. On the basis of the downfield 1H NMR spectrum in solution, His 57 is not protonated at Nepsilon2, and there is no LBHB at pH >7.0. Because addition of AcLF-CHO to chymotrypsin neither releases nor takes up a proton from solution, it is concluded that the hemiacetal oxygen of the chymotrypsin-AcLF-CHO complex is a hydroxyl group and not attracted to the oxyanion site. The protonation states of the hemiacetal and His 57 are explained by the high basicity of the hemiacetal oxygen (pK(a) > 13.5) relative to that of His 57. The 13C NMR signal for the adduct of AcLF-13CHO with chymotrypsin is consistent with a neutral hemiacetal between pH 7 and 13. At pH <7.0, His 57 in the AcLF-CHO-hemiacetal complex of chymotrypsin undergoes protonation at Nepsilon2 of His 57, leading to a transition of the 15.1 ppm downfield signal to 17.8 ppm. The pK(a)s in the active sites of the AcLF-CF3 and AcLF-CHO adducts suggest an energy barrier of 6-7 kcal x mol(-1) against ionizations that change the electrostatic charge at the active site. However, ionizations of neutral His 57 in the AcLF-CHO-chymotrypsin adduct, or in free chymotrypsin, proceed with no apparent barrier. Protonation of His 57 is accompanied by LBHB formation, suggesting that stabilization by the LBHB overcomes the barrier to ionization. On the basis of the hydration constant for AcLF-13CHO and its inhibition constant, its K(d) is 16 microM, 8000-fold larger than the comparable value for AcLF-CF3.     

Contribution of a Low-Barrier Hydrogen Bond to Catalysis Is Not Significant in Ketosteroid Isomerase

Low-barrier hydrogen bonds (LBHBs) have been proposed to have important influences on the enormous reaction rate increases achieved by many enzymes. Δ⁵-3-ketosteroid isomerase (KSI) catalyzes the allylic isomerization of Δ⁵-3-ketosteroid to its conjugated Δ⁴-isomers at a rate that approaches the diffusion limit. Tyr14, a catalytic residue of KSI, has been hypothesized to form an LBHB with the oxyanion of a dienolate steroid intermediate generated during the catalysis. The unusual chemical shift of a proton at 16.8 ppm in the nuclear magnetic resonance spectrum has been attributed to an LBHB between Tyr14 Oη and C3-O of equilenin, an intermediate analogue, in the active site of D38N KSI. This shift in the spectrum was not observed in Y30F/Y55F/D38N and Y30F/Y55F/Y115F/D38N mutant KSIs when each mutant was complexed with equilenin, suggesting that Tyr14 could not form LBHB with the intermediate analogue in these mutant KSIs. The crystal structure of Y30F/Y55F/Y115F/D38N-equilenin complex revealed that the distance between Tyr14 Oη and C3-O of the bound steroid was within a direct hydrogen bond. The conversion of LBHB to an ordinary hydrogen bond in the mutant KSI reduced the binding affinity for the steroid inhibitors by a factor of 8.1–11. In addition, the absence of LBHB reduced the catalytic activity by only a factor of 1.7–2. These results suggest that the amount of stabilization energy of the reaction intermediate provided by LBHB is small compared with that provided by an ordinary hydrogen bond in KSI.     

A low-barrier hydrogen bond between histidine of secreted phospholipase A2 and a transition state analog inhibitor

This work describes in-depth NMR characterization of a unique low-barrier hydrogen bond (LBHB) between an active site residue from the enzyme and a bound inhibitor: the complex between secreted phospholipase A(2) (sPLA(2), from bee venom and bovine pancreas) and a transition-state analog inhibitor HK32. A downfield proton NMR resonance, at 17-18 ppm, was observed in the complex but not in the free enzyme. On the basis of site-specific mutagenesis and specific 15N-decoupling, this downfield resonance was assigned to the active site H48, which is part of the catalytic dyad D99-H48. These results led to a hypothesis that the downfield resonance represents the proton (H(epsilon 2) of H48) involved in the H-bonding between D99 and H48, in analogy with serine proteases. However, this was shown not to be the case by use of the bovine enzyme labeled with specific [15N(epsilon 2)]His. Instead, the downfield resonance arises from H(delta1) of H48, which forms a hydrogen bond with a non-bridging phosphonate oxygen of the inhibitor. Further studies showed that this proton displays a fractionation factor of 0.62(+/-0.06), and an exchange rate protection factor of >100 at 285 K and >40 at 298 K, which are characteristic of a LBHB. The pK(a) of the imidazole ring of H48 was shown to be shifted from 5.7 for the free enzyme to an apparent value of 9.0 in the presence of the inhibitor. These properties are very similar to those of the Asp em leader His LBHBs in serine proteases. Possible structural bases and functional consequences for the different locations of the LBHB between these two types of enzymes are discussed. The results also underscore the importance of using specific isotope labeling, rather than extrapolation of NMR results from other enzyme systems, to assign the downfield proton resonance to a specific hydrogen bond. Although our studies did not permit the strength of the LBHB to be accurately measured, the data do not provide support for an unusually strong hydrogen bond strength (i.e. >10 kcal/mol).     

The barrier for proton transport in aquaporins as a challenge for electrostatic models: the role of protein relaxation in mutational calculations

The origin of the barrier for proton transport through the aquaporin channel is a problem of general interest. It is becoming increasingly clear that this barrier is not attributable to the orientation of the water molecules across the channel but rather to the electrostatic penalty for moving the proton charge to the center of the channel. However, the reason for the high electrostatic barrier is still rather controversial. It has been argued by some workers that the barrier is due to the so-called NPA motif and/or to the helix macrodipole or to other specific elements. However, our works indicated that the main reason for the high barrier is the loss of the generalized solvation upon moving the proton charge from the bulk to the center of the channel and that this does not reflect a specific repulsive electrostatic interaction but the absence of sufficient electrostatic stabilization. At this stage it seems that the elucidation and clarification of the origin of the electrostatic barrier can serve as an instructive test case for electrostatic models. Thus, we reexamine the free-energy surface for proton transport in aquaporins using the microscopic free-energy perturbation/umbrella sampling (FEP/US) and the empirical valence bond/umbrella sampling (EVB/US) methods as well as the semimacroscopic protein dipole Langevin dipole model in its linear response approximation version (the PDLD/S-LRA). These extensive studies help to clarify the nature of the barrier and to establish the "reduced solvation effect" as the primary source of this barrier. That is, it is found that the barrier is associated with the loss of the generalized solvation energy (which includes of course all electrostatic effects) upon moving the proton charge from the bulk solvent to the center of the channel. It is also demonstrated that the residues in the NPA region and the helix dipole cannot be considered as the main reasons for the electrostatic barrier. Furthermore, our microscopic and semimacroscopic studies clarify the problems with incomplete alternative calculations, illustrating that the effects of various electrostatic elements are drastically overestimated by macroscopic calculations that use a low dielectric constant and do not consider the protein reorganization. Similarly, it is pointed out that microscopic potential of mean force calculations that do not evaluate the electrostatic barrier relative to the bulk water cannot be used to establish the origin of the electrostatic barrier. The relationship between the present study and calculations of pK(a)s in protein interiors is clarified, pointing out that approaches that are applied to study the aquaporin barrier should be validated by pK(a)s calculations. Such calculations also help to clarify the crucial role of solvation energies in establishing the barrier in aquaporins.     

Computational and mutational analysis of human glutaredoxin (thioltransferase): probing the molecular basis of the low pKa of cysteine 22 and its role in catalysis

Human glutaredoxin (GRx), also known as thioltransferase, is a 12 kDa thiol-disulfide oxidoreductase that is highly selective for reduction of glutathione-containing mixed disulfides. The apparent pK(a) for the active site Cys22 residue is approximately 3.5. Previously we observed that the catalytic enhancement by glutaredoxin could be ascribed fully to the difference between the pK(a) of its Cys22 thiol moiety and the pK(a) of the product thiol, each acting as a leaving group in the enzymatic and nonenzymatic reactions, respectively [Srinivasan et al. (1997), Biochemistry 36, 3199-3206]. Continuum electrostatic calculations suggest that the low pK(a) of Cys22 results primarily from stabilization of the thiolate anion by a specific ion-pairing with the positively charged Lys19 residue, although hydrogen bonding interactions with Thr21 also appear to contribute. Variants of Lys19 were considered to further assess the predicted role of Lys19 on the pK(a) of Cys22. The variants K19Q and K19L were generated by molecular modeling, and the pK(a) value for Cys22 was calculated for each variant. For K19Q, the predicted Cys22 pK(a) is 7.3, while the predicted value is 8.3 for K19L. The effects of the mutations on the interaction energy between the adducted glutathionyl moiety and GRx were roughly estimated from the van der Waals and electrostatic energies between the glutathionyl moiety and proximal protein residues in a mixed disulfide adduct of GRx and glutathione, i.e., the GRx-SSG intermediate. The values for the K19 mutants differed by only a small amount compared to those for the wild type enzyme intermediate. Together, the computational analysis predicted that the mutant enzymes would have markedly reduced catalytic rates while retaining the glutathionyl specificity displayed by the wild type enzyme. Accordingly, we constructed and characterized the K19L and K19Q mutants of two forms of the GRx enzyme. Each of the mutants retained glutathionyl specificity as predicted and displayed diminution in activity, but the decreases in activity were not to the extent predicted by the theoretical calculations. Changes in the respective Cys22-thiol pK(a) values of the mutant enzymes, as shown by pH profiles for iodoacetamide inactivation of the respective enzymes, clearly revealed that the K19-C22 ion pair cannot fully account for the low pK(a) of the Cys22 thiol. Additional contributions to stabilization of the Cys22 thiolate are likely donated by Thr21 and the N-terminal partial positive charge of the neighboring alpha-helix.     

Iron-blocking the high-affinity Mn-binding site in photosystem II facilitates identification of the type of hydrogen bond participating in proton-coupled electron transport via YZ

Incubation of Mn-depleted PSII membranes [PSII(-Mn)] with Fe(II) is accompanied by the blocking of Y(Z)(*) at the high-affinity Mn-binding site to exogenous electron donors [Semin et al. (2002) Biochemistry 41, 5854-5864] and a shift of the pK(app) of the hydrogen bond partner for Y(Z) (base B) from 7.1 to 6.1 [Semin, B. K., and Seibert, M. (2004) Biochemistry 43, 6772-6782]. Here we calculate activation energies (E(a)) for Y(Z)(*) reduction in PSII(-Mn) and Fe-blocked PSII(-Mn) samples [PSII(-Mn, +Fe)] from temperature dependencies of the rate constants of the fast and slow components of the flash-probe fluorescence decay kinetics. At pH < pK(app) (e.g., 5.5), the decays are fit with one (fast) component in both types of samples, and E(a) is equal to 42.2 +/- 2.9 kJ/mol in PSII(-Mn) and 46.4 +/- 3.3 kJ/mol in PSII(-Mn, +Fe) membranes. At pH > pK(app), the decay kinetics exhibit an additional slow component in PSII(-Mn, +Fe) membranes (E(a) = 36.1 +/- 7.5 kJ/mol), which is much lower than the E(a) of the corresponding component observed for Y(Z)(*) reduction in PSII(-Mn) samples (48.1 +/- 1.7 kJ/mol). We suggest that the above difference results from the formation of a strong low barrier hydrogen bond (LBHB) between Y(Z) and base B in PSII(-Mn, +Fe) samples. To confirm this, Fe-blocking was performed in D(2)O to insert D(+), which has an energetic barrier distinct from H(+), into the LBHB. Measurement of the pH effects on the rates of Y(Z)(*) reduction in PSII(-Mn, +Fe) samples blocked in D(2)O shows a shift of the pK(app) from 6.1 to 7.6, and an increase in the E(a) of the slow component. This approach was also used to measure the stability of the Y(Z)(*) EPR signal at various temperatures in both kinds of membranes. In PSII(-Mn) membranes, the freeze-trapped Y(Z)(*) radical is stable below 190 K, but half of the Y(Z)(*) EPR signal disappears after a 1-min incubation when the sample is warmed to 253 K. In PSII(-Mn, +Fe) samples, the trapped Y(Z)(*) radical is unstable at a much lower temperature (77 K). However, the insertion of D(+) into the hydrogen bond between Y(Z) and base B during the blocking process increases the temperature stability of the Y(Z)(*) EPR signal at 77 K. Again, these results indicate that Fe-blocking involves Y(Z) in the formation of a LBHB, which in turn is consistent with the suggested existence of a LBHB between Y(Z) and base B in intact PSII membranes [Zhang, C., and Styring, S. (2003) Biochemistry 42, 8066-8076].     

A novel engineered subtilisin BPN' lacking a low-barrier hydrogen bond in the catalytic triad

The low-barrier hydrogen bond (LBHB) between the Asp and His residues of the catalytic triad in a serine protease was perturbed via the D32C mutation in subtilisin BPN' (Bacillus protease N'). This mutant enzyme catalyzes the hydrolysis of N-Suc-Ala-Ala-Pro-Phe-SBzl with a k(cat)/K(m) value that is only 8-fold reduced from that of the wild-type (WT) enzyme. The value of k(cat)/K(m) for the corresponding p-nitroanilide (pNA) substrate is only 50-fold lower than that of the WT enzyme (DeltaDeltaG++ = 2.2 kcal/mol). The pK(a) controlling the ascending limb of the pH versus k(cat)/K(m) profile is lowered from 7.01 (WT) to 6.53 (D32C), implying that any hydrogen bond replacing that between Asp32 and His64 of the WT enzyme most likely involves the neutral thiol rather than the thiolate form of Cys32. It is shown by viscosity variation that the reaction of WT subtilisin with N-Suc-Ala-Ala-Pro-Phe-SBzl is 50% (sucrose) to 100% (glycerol) diffusion-controlled, while that of the D32C construct is 29% (sucrose) to 76% (glycerol) diffusion-controlled. The low-field NMR resonance of 18 ppm that has been assigned to a proton shared by Asp32 and His64, and is considered diagnostic of a LBHB in the WT enzyme, is not present in D32C subtilisin. Thus, the LBHB is not an inherent requirement for substantial rate enhancement for subtilisin.     

Assessing the quality of the homology-modeled 3D structures from electrostatic standpoint: test on bacterial nucleoside monophosphate kinase families

In this study, we address the issue of performing meaningful pK(a) calculations using homology modeled three-dimensional (3D) structures and analyze the possibility of using the calculated pK(a) values to detect structural defects in the models. For this purpose, the 3D structure of each member of five large protein families of a bacterial nucleoside monophosphate kinases (NMPK) have been modeled by means of homology-based approach. Further, we performed pK(a) calculations for the each model and for the template X-ray structures. Each bacterial NMPK family used in the study comprised on average 100 members providing a pool of sequences and 3D models large enough for reliable statistical analysis. It was shown that pK(a) values of titratable groups, which are highly conserved within a family, tend to be conserved among the models too. We demonstrated that homology modeled structures with sequence identity larger than 35% and gap percentile smaller than 10% can be used for meaningful pK(a) calculations. In addition, it was found that some highly conserved titratable groups either exhibit large pK(a) fluctuations among the models or have pK(a) values shifted by several pH units with respect to the pK(a) calculated for the X-ray structure. We demonstrated that such case usually indicates structural errors associated with the model. Thus, we argue that pK(a) calculations can be used for assessing the quality of the 3D models by monitoring fluctuations of the pK(a) values for highly conserved titratable residues within large sets of homologous proteins.     

Salt effects on ionization equilibria of histidines in myoglobin

The salt dependence of histidine pK(a) values in sperm whale and horse myoglobin and in histidine-containing peptides was measured by (1)H-NMR spectroscopy. Structure-based pK(a) calculations were performed with continuum methods to test their ability to capture the effects of solution conditions on pK(a) values. The measured pK(a) of most histidines, whether in the protein or in model compounds, increased by 0.3 pH units or more between 0.02 M and 1.5 M NaCl. In myoglobin two histidines (His(48) and His(36)) exhibited a shallower dependence than the average, and one (His(113)) showed a steeper dependence. The (1)H-NMR data suggested that the salt dependence of histidine pK(a) values in the protein was determined primarily by the preferential stabilization of the charged form of histidine with increasing salt concentrations rather than by screening of electrostatic interactions. The magnitude and salt dependence of interactions between ionizable groups were exaggerated in pK(a) calculations with the finite-difference Poisson-Boltzmann method applied to a static structure, even when the protein interior was treated with arbitrarily high dielectric constants. Improvements in continuum methods for calculating salt effects on pK(a) values will require explicit consideration of the salt dependence of model compound pK(a) values used for reference in the calculations.     

Role of active site residues in the glycosylase step of T4 endonuclease V. Computer simulation studies on ionization states.

T4 Endonuclease V (EndoV) is a base excision repair enzyme that removes thymine dimers (TD) from damaged DNA. To elucidate the role of the active site residues in catalysis, their pK(a)'s were evaluated using the semimicroscopic version of the protein dipoles-Langevin dipoles method (PDLD/S). Contributions of different effects to the pK(a) such as charge-charge interactions, conformational rearrangement, protein relaxation, and DNA binding were analyzed in detail. Charging of the active site residues was found to be less favorable in the complex than in the free enzyme. The pK(a) of the N-terminus decreased from 8.01 in the free enzyme to 6.52 in the complex, while the pK(a) of Glu-23 increased from 1. 52 to 7.82, which indicates that the key residues are neutral in the reactant state of the glycosylase step. These pK(a)'s are in agreement with the optimal pH range of the reaction and support the N-terminus acting as a nucleophile. The Glu-23 in its protonated form is hydrogen bonded to O4' of the sugar of 5' TD and can play a role in increasing the positive charge of C1' and, hence, accelerating the nucleophilic substitution. Furthermore, the neutral Glu-23 is a likely candidate to protonate O4' to induce ring opening required to complete the glycosylase step of EndoV. The positively charged Arg-22 and Arg-26 provide an electrostatically favorable environment for the leaving base. To distinguish between S(N)1 and S(N)2 mechanisms of the glycosylase step the energetics of protonating O2 of 5' TD was calculated. The enzyme was found to stabilize the neutral thymine by approximately 3.6 kcal/mol, whereas it destabilizes the protonated thymine by approximately 6.6 kcal/mol with respect to an aqueous environment. Consequently, the formation of a protonated thymine intermediate is unlikely, indicating an S(N)2 reaction mechanism for the glycosylase step.     

Structure, dynamics and electrostatics of the active site of glutaredoxin 3 from Escherichia coli: comparison with functionally related proteins.

The chemistry of active-site cysteine residues is central to the activity of thiol-disulfide oxidoreductases of the thioredoxin superfamily. In these reactions, a nucleophilic thiolate is required, but the associated pK(a) values differ vastly in the superfamily, from less than 4 in DsbA to greater than 7 in Trx. The factors that stabilize this thiolate are, however, not clearly established. The glutaredoxins (Grxs), which are members of this superfamily, contain a Cys-Pro-Tyr-Cys motif in their active site. In reduced Grxs, the pK(a) of the N-terminal active-site nucleophilic cysteine residue is lowered significantly, and the stabilization of the corresponding thiolate is expected to influence the redox potential of these enzymes. Here, we use a combination of long molecular dynamics (MD) simulations, pK(a) calculations, and experimental investigations to derive the structure and dynamics of the reduced active site from Escherichia coli Grx3, and investigate the factors that stabilize the thiolate. Several different MD simulations converged toward a consensus conformation for the active-site cysteine residues (Cys11 and Cys14), after a number of local conformational changes. Key features of the model were tested experimentally by measurement of NMR scalar coupling constants, and determination of pK(a) values of selected residues. The pK(a) values of the Grx3 active-site residues were calculated during the MD simulations, and support the underlying structural model. The structure of Grx3, in combination with the pK(a) calculations, indicate that the pK(a) of the N-terminal active-site cysteine residue in Grx3 is intermediate between that of its counterpart in DsbA and Trx. The pK(a) values in best agreement with experiment are obtained with a low (<4) protein dielectric constant. The calculated pK(a) values fluctuate significantly in response to protein dynamics, which underscores the importance of the details of the underlying structures when calculating pK(a) values. The thiolate of Cys11 is stabilized primarily by direct hydrogen bonding with the amide protons of Tyr13 and Cys14 and the thiol proton of Cys14, rather than by long-range interactions from charged groups or from a helix macrodipole. From the comparison of reduced Grx3 with other members of the thioredoxin superfamily, a unifying theme for the structural basis of thiol pK(a) differences in this superfamily begins to emerge.     

Detection of large pKa perturbations of an inhibitor and a catalytic group at an enzyme active site, a mechanistic basis for catalytic power of many enzymes

Delta(5)-3-Ketosteroid isomerase catalyzes cleavage and formation of a C-H bond at a diffusion-controlled limit. By determining the crystal structures of the enzyme in complex with each of three different inhibitors and by nuclear magnetic resonance (NMR) spectroscopic investigation, we evidenced the ionization of a hydroxyl group (pK(a) approximately 16.5) of an inhibitor, which forms a low barrier hydrogen bond (LBHB) with a catalytic residue Tyr(14) (pK(a) approximately 11.5), and the protonation of the catalytic residue Asp(38) with pK(a) of approximately 4.5 at pH 6.7 in the interaction with a carboxylate group of an inhibitor. The perturbation of the pK(a) values in both cases arises from the formation of favorable interactions between inhibitors and catalytic residues. The results indicate that the pK(a) difference between catalytic residue and substrate can be significantly reduced in the active site environment as a result of the formation of energetically favorable interactions during the course of enzyme reactions. The reduction in the pK(a) difference should facilitate the abstraction of a proton and thereby eliminate a large fraction of activation energy in general acid/base enzyme reactions. The pK(a) perturbation provides a mechanistic ground for the fast reactivity of many enzymes and for the understanding of how some enzymes are able to extract a proton from a C-H group with a pK(a) value as high as approximately 30.     

Optimizing the hydrogen-bond network in Poisson-Boltzmann equation-based pK(a) calculations

pK(a) calculation methods that are based on finite difference solutions to the Poisson-Boltzmann equation (FDPB) require that energy calculations be performed for a large number of different protonation states of the protein. Normally, the differences between these protonation states are modeled by changing the charges on a few atoms, sometimes the differences are modeled by adding or removing hydrogens, and in a few cases the positions of these hydrogens are optimized locally. We present an FDPB-based pK(a) calculation method in which the hydrogen-bond network is globally optimized for every single protonation state used. This global optimization gives a significant improvement in the accuracy of calculated pK(a) values, especially for buried residues. It is also shown that large errors in calculated pK(a) values are often due to structural artifacts induced by crystal packing. Optimization of the force fields and parameters used in pK(a) calculations should therefore be performed with X-ray structures that are corrected for crystal artifacts.     

High apparent dielectric constants in the interior of a protein reflect water penetration

A glutamic acid was buried in the hydrophobic core of staphylococcal nuclease by replacement of Val-66. Its pK(a) was measured with equilibrium thermodynamic methods. It was 4.3 units higher than the pK(a) of Glu in water. This increase was comparable to the DeltapK(a) of 4.9 units measured previously for a lysine buried at the same location. According to the Born formalism these DeltapK(a) are energetically equivalent to the transfer of a charged group from water to a medium of dielectric constant of 12. In contrast, the static dielectric constants of dry protein powders range from 2 to 4. In the crystallographic structure of the V66E mutant, a chain of water molecules was seen that hydrates the buried Glu-66 and links it with bulk solvent. The buried water molecules have never previously been detected in >20 structures of nuclease. The structure and the measured energetics constitute compelling and unprecedented experimental evidence that solvent penetration can contribute significantly to the high apparent polarizability inside proteins. To improve structure-based calculations of electrostatic effects with continuum methods, it will be necessary to learn to account quantitatively for the contributions by solvent penetration to dielectric effects in the protein interior.     

Low-barrier hydrogen bond hypothesis in the catalytic triad residue of serine proteases: correlation between structural rearrangement and chemical shifts in the acylation process

To elucidate the catalytic advantage of the low-barrier hydrogen bond (LBHB), we analyze the hydrogen bonding network of the catalytic triad (His57-Asp102-Ser195) of serine protease trypsin, one of the best examples of the LBHB reaction mechanism. Especially, we focus on the correlation between the change of the chemical shifts and the structural rearrangement of the active site in the acylation process. To clarify LBHB, we evaluate the two complementary properties. First, we calculate the NMR chemical shifts of the imidazole ring of His57 by the gauge-including atomic orbital (GIAO) approach within the ab initio QM/MM framework. Second, the free energy profile of the proton transfer from His57 to Asp102 in the tetrahedral intermediate is obtained by ab initio QM/MM calculations combined with molecular dynamics free energy perturbation (MD-FEP) simulations. The present analyses reveal that the calculated shifts reasonably reproduce the observed values for (1)H chemical shift of H(epsilon)(1) and H(delta)(1) in His57. The (15)N and (13)C chemical shifts are also consistent with the experiments. It is also shown that the proton between His57 and Asp102 is localized at the His57 side. This largely downfield chemical shift is originated from the strong electrostatic interaction, not a covalent-like bonding character between His57 and Asp102. Also, it is proved that a slight downfield character of H(epsilon)(1) is originated from a electrostatic interaction between His57 and the backbone carbonyl group of Val213 and Ser214. These downfield chemical shifts are observed only when the tetrahedral intermediate is formed in the acylation process.     

Evaluation of linked protonation effects in protein binding reactions using isothermal titration calorimetry.

A theoretical development in the evaluation of proton linkage in protein binding reactions by isothermal titration calorimetry (ITC) is presented. For a system in which binding is linked to protonation of an ionizable group on a protein, we show that by performing experiments as a function of pH in buffers with varying ionization enthalpy, one can determine the pK(a)'s of the group responsible for the proton linkage in the free and the liganded states, the protonation enthalpy for this group in these states, as well as the intrinsic energetics for ligand binding (delta H(o), delta S(o), and delta C(p)). Determination of intrinsic energetics in this fashion allows for comparison with energetics calculated empirically from structural information. It is shown that in addition to variation of the ligand binding constant with pH, the observed binding enthalpy and heat capacity change can undergo extreme deviations from their intrinsic values, depending upon pH and buffer conditions.     

H atom positions and nuclear magnetic resonance chemical shifts of short H bonds in photoactive yellow protein

Recent neutron diffraction studies on photoactive yellow protein (PYP) proposed that the H bond between protonated Glu46 and the chromophore-ionized p-coumaric acid (pCA) is a low-barrier H bond (LBHB) mainly because the H atom position was assigned at the midpoint of the O(Glu46)-O(pCA) bond. However, the (1)H nuclear magnetic resonance (NMR) chemical shift (δ(H)) was 15.2 ppm, which is lower than the values of 17-19 ppm for typical LBHBs. We evaluated the dependence of δ(H) on an H atom position in the O(Glu46)-O(pCA) bond in the PYP ground state by using a quantum mechanical/molecular mechanical (QM/MM) approach. The calculated chemical shift unambiguously suggested that a δ(H) of 15.2 ppm for the O(Glu46)-O(pCA) bond in NMR studies should correspond to the QM/MM geometry (δ(H) = 14.5 ppm), where the H atom belongs to the Glu moiety, rather than the neutron diffraction geometry (δ(H) = 19.7 ppm), where the H atom is near the midpoint of the donor and acceptor atoms.