Evaluation of protein-protein association energies by free energy perturbation calculations

The association energy upon binding of different amino acids in the specificity pocket of trypsin was evaluated by free energy perturbation calculations on complexes between bovine trypsin (BT) and bovine pancreatic trypsin inhibitor (BPTI). Three simulations of mutations of the primary binding residue (P(1)) were performed (P(1)-Ala to Gly, P(1)-Met to Gly and P(1)-Met to Ala) and the resulting differences in association energy (DeltaDeltaG(a)) are 2. 28, 5.08 and 2.93 kcal/mol for P(1)-Ala to Gly, P(1)-Met to Gly and to Ala with experimental values of 1.71, 4.62 and 2.91 kcal/mol, respectively. The calculated binding free energy differences are hence in excellent agreement with the experimental binding free energies. The binding free energies, however, were shown to be highly dependent on water molecules at the protein-protein interface and could only be quantitatively estimated if the correct number of such water molecules was included. Furthermore, the cavities that were formed when a large amino acid side-chain is perturbed to a smaller one seem to create instabilities in the systems and had to be refilled with water molecules in order to obtain reliable results. In addition, if the protein atoms that were perturbed away were not replaced by water molecules, the simulations dramatically overestimated the initial state of the free energy perturbations.     

Free energy calculations show that acidic P1 variants undergo large pKa shifts upon binding to trypsin

Serine proteinases and their protein inhibitors belong to one of the most comprehensively studied models of protein-protein interactions. It is well established that the narrow trypsin specificity is caused by the presence of a negatively charged aspartate at the specificity pocket. X-ray crystallography as well as association measurements revealed, surprisingly, that BPTI with glutamatic acid as the primary binding (P1) residue was able to bind to trypsin. Previous free energy calculations showed that there was a substantially unfavorable binding free energy associated with accommodation of ionized P1 Glu at the S1-site of trypsin. In this study, the binding of P1 Glu to trypsin has been systematically investigated in terms of the protonation states of P1 Glu and Asp189, the orientation of Gln192, as well as the possible presence of counterions using the linear interaction energy (LIE) approach and the free energy perturbation (FEP) method. Twenty-four conceivable binding arrangements were evaluated and quantitative agreement with experiments is obtained when the P1 Glu binds in its protonated from. The results suggest that P1 Glu is one of the variants of BPTI that inhibit trypsin strongest at low pH, contrary to the specificity profile of trypsin, suggesting a new regulation mechanism of trypsin-like enzymes.     

Structure-function relationship of serine protease-protein inhibitor interaction.

We report our progress in understanding the structure-function relationship of the interaction between protein inhibitors and several serine proteases. Recently, we have determined high resolution solution structures of two inhibitors Apis mellifera chymotrypsin inhibitor-1 (AMCI-I) and Linum usitatissimum trypsin inhibitor (LUTI) in the free state and an ultra high resolution X-ray structure of BPTI. All three inhibitors, despite totally different scaffolds, contain a solvent exposed loop of similar conformation which is highly complementary to the enzyme active site. Isothermal calo- rimetry data show that the interaction between wild type BPTI and chymotrypsin is entropy driven and that the enthalpy component opposes complex formation. Our research is focused on extensive mutagenesis of the four positions from the protease binding loop of BPTI: P1, P1', P3, and P4. We mutated these residues to different amino acids and the variants were characterized by determination of the association constants, stability parameters and crystal structures of protease-inhibitor complexes. Accommodation of the P1 residue in the S1 pocket of four proteases: chymotrypsin, trypsin, neutrophil elastase and cathepsin G was probed with 18 P1 variants. High resolution X-ray structures of ten complexes between bovine trypsin and P1 variants of BPTI have been determined and compared with the cognate P1 Lys side chain. Mutations of the wild type Ala16 (P1') to larger side chains always caused a drop of the association constant. According to the crystal structure of the Leu16 BPTI-trypsin complex, introduction of the larger residue at the P1' position leads to steric conflicts in the vicinity of the mutation. Finally, mutations at the P4 site allowed an improvement of the association with several serine proteases involved in blood clotting. Conversely, introduction of Ser, Val, and Phe in place of Gly12 (P4) had invariably a destabilizing effect on the complex with these proteases.     

Probing the effect of point mutations at protein-protein interfaces with free energy calculations

We have studied the effect of point mutations of the primary binding residue (P1) at the protein-protein interface in complexes of chymotrypsin and elastase with the third domain of the turkey ovomucoid inhibitor and in trypsin with the bovine pancreatic trypsin inhibitor, using molecular dynamics simulations combined with the linear interaction energy (LIE) approach. A total of 56 mutants have been constructed and docked into their host proteins. The free energy of binding could be reliably calculated for 52 of these mutants that could unambiguously be fitted into the binding sites. We find that the predicted binding free energies are in very good agreement with experimental data with mean unsigned errors between 0.50 and 1.03 kcal/mol. It is also evident that the standard LIE model used to study small drug-like ligand binding to proteins is not suitable for protein-protein interactions. Three different LIE models were therefore tested for each of the series of protein-protein complexes included, and the best models for each system turn out to be very similar. The difference in parameterization between small drug-like compounds and protein point mutations is attributed to the preorganization of the binding surface. Our results clearly demonstrate the potential of free energy calculations for probing the effect of point mutations at protein-protein interfaces and for exploring the principles of specificity of hot spots at the interface.     

Analysis of serine proteinase-inhibitor interaction by alanine shaving

We analyzed the energetic importance of residues surrounding the hot spot (the P(1) position) of bovine pancreatic trypsin inhibitor (BPTI) in interaction with two proteinases, trypsin and chymotrypsin, by a procedure called molecular shaving. One to eight residues of the structural epitope, composed of two extended and exposed loops, were mutated to alanine(s). Although truncation of the side chains of residues surrounding the P(1) position to methyl groups caused a decrease in Delta G(den) values up to 6.4 kcal mole(-1), it did not influence the overall conformation of the inhibitor. We found that the replacement of up to six residues with alanines was fully additive at the level of protein stability. To analyze the influence of the structural epitope on the association energy, we determined association constants for BPTI variants and both enzymes and applied the additivity analysis. Shaving of two binding loops led to a progressive drop in the association energy, more pronounced for trypsin (decrease up to 9.6 kcal mole(-1)) than chymotrypsin (decrease up to 3.5 kcal mole(-1)). In the case of extensively mutated variants interacting with chymotrypsin, the association energies agreed very well with the values calculated from single mutational effects. However, when P(1)-neighboring residues were shaved to alanine(s), their contribution to the association energy was not fully removed because of the presence of methyl groups and main chain-main chain intermolecular hydrogen bonds. Moreover, the hot spot had a different contribution to the complex stability in the fully shaved BPTI variant compared with the wild type, which was caused by perturbations of the P(1)-S(1) electrostatic interaction.     

Engineering trypsin for inhibitor resistance

The development of effective protease therapeutics requires that the proteases be more resistant to naturally occurring inhibitors while maintaining catalytic activity. A key step in developing inhibitor resistance is the identification of key residues in protease-inhibitor interaction. Given that majority of the protease therapeutics currently in use are trypsin-fold, trypsin itself serves as an ideal model for studying protease-inhibitor interaction. To test the importance of several trypsin-inhibitor interactions on the prime-side binding interface, we created four trypsin single variants Y39A, Y39F, K60A, and K60V and report biochemical sensitivity against bovine pancreatic trypsin inhibitor (BPTI) and M84R ecotin. All variants retained catalytic activity against small, commercially available peptide substrates [k_cat/K_M = (1.2 ± 0.3) × 10⁷ M⁻¹ s⁻¹. Compared with wild-type, the K60A and K60V variants showed increased sensitivity to BPTI but less sensitivity to ecotin. The Y39A variant was less sensitive to BPTI and ecotin while the Y39F variant was more sensitive to both. The relative binding free energies between BPTI complexes with WT, Y39F, and Y39A were calculated based on 3.5 µs combined explicit solvent molecular dynamics simulations. The BPTI:Y39F complex resulted in the lowest binding energy, while BPTI:Y39A resulted in the highest. Simulations of Y39F revealed increased conformational rearrangement of F39, which allowed formation of a new hydrogen bond between BPTI R17 and H40 of the variant. All together, these data suggest that positions 39 and 60 are key for inhibitor binding to trypsin, and likely more trypsin-fold proteases.     

Tyrosine Sulfation of Human Trypsin Steers S2’ Subsite Selectivity towards Basic Amino Acids

Human cationic and anionic trypsins are sulfated on Tyr154, a residue which helps to shape the prime side substrate-binding subsites. Here, we used phage display technology to assess the significance of tyrosine sulfation for the specificity of human trypsins. The prime side residues P1′–P4′ in the binding loop of bovine pancreatic trypsin inhibitor (BPTI) were fully randomized and tight binding inhibitor phages were selected against non-sulfated and sulfated human cationic trypsin. The selection pattern for the two targets differed mostly at the P2′ position, where variants selected against non-sulfated trypsin contained primarily aliphatic residues (Leu, Ile, Met), while variants selected against sulfated trypsin were enriched also for Arg. BPTI variants carrying Arg, Lys, Ile, Leu or Ala at the P2′ position of the binding loop were purified and equilibrium dissociation constants were determined against non-sulfated and sulfated cationic and anionic human trypsins. BPTI variants harboring apolar residues at P2′ exhibited 3–12-fold lower affinity to sulfated trypsin relative to the non-sulfated enzyme, whereas BPTI variants containing basic residues at P2′ had comparable affinity to both trypsin forms. Taken together, the observations demonstrate that the tyrosyl sulfate in human trypsins interacts with the P2′ position of the substrate-like inhibitor and this modification increases P2′ selectivity towards basic side chains.     

Energetics of codon-anticodon recognition on the small ribosomal subunit

Recent crystal structures of the small ribosomal subunit have made it possible to examine the detailed energetics of codon recognition on the ribosome by computational methods. The binding of cognate and near-cognate anticodon stem loops to the ribosome decoding center, with mRNA containing the Phe UUU and UUC codons, are analyzed here using explicit solvent molecular dynamics simulations together with the linear interaction energy (LIE) method. The calculated binding free energies are in excellent agreement with experimental binding constants and reproduce the relative effects of mismatches in the first and second codon position versus a mismatch at the wobble position. The simulations further predict that the Leu2 anticodon stem loop is about 10 times more stable than the Ser stem loop in complex with the Phe UUU codon. It is also found that the ribosome significantly enhances the intrinsic stability differences of codon-anticodon complexes in aqueous solution. Structural analysis of the simulations confirms the previously suggested importance of the universally conserved nucleotides A1492, A1493, and G530 in the decoding process.     

Crystal structures of five bovine chymotrypsin complexes with P1 BPTI variants

The bovine chymotrypsin-bovine pancreatic trypsin inhibitor (BPTI) interaction belongs to extensively studied models of protein-protein recognition. The accommodation of the inhibitor P1 residue in the S1 binding site of the enzyme forms the hot spot of this interaction. Mutations introduced at the P1 position of BPTI result in a more than five orders of magnitude difference of the association constant values with the protease. To elucidate the structural aspects of the discrimination between different P1 residues, crystal structures of five bovine chymotrypsin-P1 BPTI variant complexes have been determined at pH 7.8 to a resolution below 2 A. The set includes polar (Thr), ionizable (Glu, His), medium-sized aliphatic (Met) and large aromatic (Trp) P1 residues and complements our earlier studies of the interaction of different P1 side-chains with the S1 pocket of chymotrypsin. The structures have been compared to the complexes of proteases with similar and dissimilar P1 preferences, including Streptomyces griseus proteases B and E, human neutrophil elastase, crab collagenase, bovine trypsin and human thrombin. The S1 sites of these enzymes share a common general shape of significant rigidity. Large and branched P1 residues adapt in their complexes similar conformations regardless of the polarity and size differences between their S1 pockets. Conversely, long and flexible residues such as P1 Met are present in the disordered form and display a conformational diversity despite similar inhibitory properties with respect to most enzymes studied. Thus, the S1 specificity profiles of the serine proteases appear to result from the precise complementarity of the P1-S1 interface and minor conformational adjustments occurring upon the inhibitor binding.     

Interscaffolding additivity: binding of P1 variants of bovine pancreatic trypsin inhibitor to four serine proteases.

Different families of protein inhibitors of serine proteases share similar conformation of the enzyme-binding loop, while their scaffoldings are completely different. In the enzyme-inhibitor complex, the P1position of the loop makes numerous contacts within the S1pocket and significantly influences the energy of the interaction. Here, we determine the association energies (DeltaGavalues) for the interaction of coded P1variants of bovine pancreatic trypsin inhibitor (BPTI) with bovine beta-trypsin (BT), anionic salmon trypsin (AST), bovine alpha-chymotrypsin (BCHYM), and human neutrophil elastase (HNE). The respective DeltaGaranges are 15, 13, 9, and 8 kcal mol-1(1 cal=4.18 J). Next, through interscaffolding additivity cycles, we compare our set of DeltaGavalues determined for BCHYM and HNE with similar data sets available in the literature for three other inhibitor families. The analysis of the cycles shows that 27 to 83 % of cycles fulfil the criteria of additvity. In one particular case (comparison of associations of P1variants of BPTI and OMTKY3 with BCHYM) there is a structural basis for strongly non-additive behaviour. We argue that the interscaffolding additvity depends on sequential and conformational similarities of sites where the mutation(s) are introduced and on the particular substitution. In the second interscaffolding analysis, we compare binding of the same P1mutants to BT and AST. The high correlation coefficient shows that both trypsins recognize with comparable strength the non-cognate side-chains. However, the cognate Arg and Lys side-chains are recognized significantly more strongly by AST.     

Structural consequences of accommodation of four non-cognate amino acid residues in the S1 pocket of bovine trypsin and chymotrypsin

Crystal structures of P1 Gly, Val, Leu and Phe bovine pancreatic trypsin inhibitor (BPTI) variants in complex with two serine proteinases, bovine trypsin and chymotrypsin, have been determined. The association constants for the four mutants with the two enzymes show that the enlargement of the volume of the P1 residue is accompanied by an increase of the binding energy, which is more pronounced for bovine chymotrypsin. Since the conformation of the P1 side-chains in the two S1 pockets is very similar, we suggest that the difference in DeltaG values between the enzymes must arise from the more polar environment of the S1 site of trypsin. This results mainly from the substitutions of Met192 and Ser189 observed in chymotrypsin with Gln192 and Asp189 present in trypsin. The more polar interior of the S1 site of trypsin is reflected by a much higher order of the solvent network in the empty pocket of the enzyme, as is observed in the complexes of the two enzymes with the P1 Gly BPTI variant. The more optimal binding of the large hydrophobic P1 residues by chymotrypsin is also reflected by shrinkage of the S1 pocket upon the accommodation of the cognate residues of this enzyme. Conversely, the S1 pocket of trypsin expands upon binding of such side-chains, possibly to avoid interaction with the polar residues of the walls. Further differentiation between the two enzymes is achieved by small differences in the shape of the S1 sites, resulting in an unequal steric hindrance of some of the side-chains, as observed for the gamma-branched P1 Leu variant of BPTI, which is much more favored by bovine chymotrypsin than trypsin. Analysis of the discrimination of beta-branched residues by trypsin and chymotrypsin is based on the complexes with the P1 Val BPTI variant. Steric repulsion of the P1 Val residue by the walls of the S1 pocket of both enzymes prevents the P1 Val side-chain from adopting the most optimal chi1 value.     

Analysis of cyclosporin A and a set of analogs as inhibitors of a T. cruzi cyclophilin by docking and molecular dynamics

Cyclophilins (CyPs) are enzymes involved in protein folding. In Trypanosoma cruzi (T. cruzi), the most abundantly expressed CyP is the isoform TcCyP19. It has been shown that TcCyP19 is inhibited by the immunosuppressive drug cyclosporin A (CsA) and analogs, which also proved to have potent trypanosomicidal activity in vitro. In this work, we continue and expand a previous study on the molecular interactions of CsA, and a set of analogs modeled in complexes with TcCyP19. The modeled complexes were used to evaluate binding free energies by molecular dynamics (MD), applying the Linear Interaction Energy (LIE) method. In addition, putative binding sites were identified by molecular docking. In our analysis, the binding free energy calculations did not correlate with experimental data. The heterogeneity of the non-bonded energies and the variation in the pattern of hydrogen bonds suggest that the systems may not be suitable for the application of the LIE method. Further, the docking calculations identified two other putative binding sites with comparable scoring energies to the active site, a fact that may also explain the lack of correlation found. Kinetic experiments are needed to confirm or reject the multiple binding sites hypothesis. In the meantime, MD simulations at the alternative sites, employing other methods to compute binding free energies, might be successful at finding good correlations with the experimental data.     

The crystal structures of the complexes between bovine beta-trypsin and ten P1 variants of BPTI

The high-resolution X-ray structures have been determined for ten complexes formed between bovine beta-trypsin and P1 variants (Gly, Asp, Glu, Gln, Thr, Met, Lys, His, Phe, Trp) of bovine pancreatic trypsin inhibitor (BPTI). All the complexes were crystallised from the same conditions. The structures of the P1 variants Asp, Glu, Gln and Thr, are reported here for the first time in complex with any serine proteinase. The resolution of the structures ranged from 1.75 to 2.05 A and the R-factors were about 19-20 %. The association constants of the mutants ranged from 1.5x10(4) to 1.7x10(13) M-1. All the structures could be fitted into well-defined electron density, and all had very similar global conformations. All the P1 mutant side-chains could be accomodated at the primary binding site, but relative to the P1 Lys, there were small local changes within the P1-S1 interaction site. These comprised: (1) changes in the number and dynamics of water molecules inside the pocket; (2) multiple conformations and non-optimal dihedral angles for some of the P1 side-chains, Ser190 and Gln192; and (3) changes in temperature factors of the pocket walls as well as the introduced P1 side-chain. Binding of the cognate P1 Lys is characterised by almost optimal dihedral angles, hydrogen bonding distances and angles, in addition to considerably lower temperature factors. Thus, the trypsin S1 pocket seems to be designed particularly for lysine binding.     

The pancreatic islet β-cell-enriched transcription factor Pdx-1 regulates Slc30a8 gene transcription through an intronic enhancer

PRSS3/mesotrypsin is an atypical isoform of trypsin, the up-regulation of which has been implicated in promoting tumour progression. Mesotrypsin inhibitors could potentially provide valuable research tools and novel therapeutics, but small-molecule trypsin inhibitors have low affinity and little selectivity, whereas protein trypsin inhibitors bind poorly and are rapidly degraded by mesotrypsin. In the present study, we use mutagenesis of a mesotrypsin substrate, APPI (amyloid precursor protein Kunitz protease inhibitor domain), and of a poor mesotrypsin inhibitor, BPTI (bovine pancreatic trypsin inhibitor), to dissect mesotrypsin specificity at the key P(2)' position. We find that bulky and charged residues strongly disfavour binding, whereas acidic residues facilitate catalysis. Crystal structures of mesotrypsin complexes with BPTI variants provide structural insights into mesotrypsin specificity and inhibition. Through optimization of the P(1) and P(2)' residues of BPTI, we generate a stable high-affinity mesotrypsin inhibitor with an equilibrium binding constant K(i) of 5.9 nM, a >2000-fold improvement in affinity over native BPTI. Using this engineered inhibitor, we demonstrate the efficacy of pharmacological inhibition of mesotrypsin in assays of breast cancer cell malignant growth and pancreatic cancer cell invasion. Although further improvements in inhibitor selectivity will be important before clinical potential can be realized, the results of the present study support the feasibility of engineering protein protease inhibitors of mesotrypsin and highlight their therapeutic potential.     

Phage display selection of P1 mutants of BPTI directed against five different serine proteinases

The P1 position of protein inhibitors and oligopeptide substrates determines, to a large extent, association energy with many serine proteinases. To test the agreement of phage display selection with the existing thermodynamic data, a small library of all 20 P1 mutants of basic pancreatic trypsin inhibitor (BPTI) was created, fused to protein III, and displayed on the surface of M13 phage. The wild type of displayed inhibitor monovalently and strongly inhibited trypsin with an association constant of Ka = 3 x 10(11) M(-1). The library was applied to select BPTI variants active against five serine proteinases of different specificity (bovine trypsin and chymotrypsin, human leukocyte and porcine pancreatic elastases, human azurocidin). The results of enrichment with four proteinases agreed well with the available thermodynamic data. In the case of azurocidin, the phage display selection allowed determination of the P1 specificity of this protein with the following frequencies for selected P1 variants: 43% Lys, 36% Leu, 7% Met, 7% Thr, 7% Gln.     

Electrostatic effects play a central role in cold adaptation of trypsin.

Organisms that live in constantly cold environments have to adapt their metabolism to low temperatures, but mechanisms of enzymatic adaptation to cold environments are not fully understood. Cold active trypsin catalyses reactions more efficiently and binds ligands more strongly in comparison to warm active trypsin. We have addressed this issue by means of comparative free energy calculations studying the binding of positively charged ligands to two trypsin homologues. Stronger inhibition of the cold active trypsin by benzamidine and positively charged P1-variants of BPTI is caused by rather subtle electrostatic effects. The different affinity of benzamidine originates solely from long range interactions, while the increased binding of P1-Lys and -Arg variants of BPTI is attributed to both long and short range effects that are enhanced in the cold active trypsin compared to the warm active counterpart. Electrostatic interactions thus provide an efficient strategy for cold adaptation of trypsin.     

Substitutions at the P(1) position in BPTI strongly affect the association energy with serine proteinases

The role of the S(1) subsite in trypsin, chymotrypsin and plasmin has been examined by measuring the association with seven different mutants of bovine pancreatic trypsin inhibitor (BPTI); the mutants contain Gly, Ala, Ser, Val, Leu, Arg, and Trp at the P(1) position of the reactive site. The effects of substitutions at the P(1) position on the association constants are very large, comprising seven orders of magnitude for trypsin and plasmin, and over five orders for chymotrypsin. All mutants showed a decrease of the association constant to the three proteinases in the same order: Ala>Gly>Ser>Arg>Val>Leu>Trp. Calorimetric and circular dichroism methods showed that none of the P1 substitutions, except the P1-Val mutant, lead to destabilisation of the binding loop conformation. The X-ray structure of the complex formed between bovine beta-trypsin and P(1)-Leu BPTI showed that the P(1)-Leu sterically conflicts with the side-chain of P(3)-Ile, which thereby is forced to rotate approximately 90 degrees. Ile18 (P(3)) in its new orientation, in turn interacts with the Tyr39 side-chain of trypsin. Introduction of a large side-chain at the P1' position apparently leads to a cascade of small alterations of the trypsin-BPTI interface that seem to destabilise the complex by it adopting a less optimized packing and by tilting the BPTI molecule up to 15 degrees compared to the native trypsin-BPTI complex.     

Improving the iterative Linear Interaction Energy approach using automated recognition of configurational transitions

Recently an iterative method was proposed to enhance the accuracy and efficiency of ligand-protein binding affinity prediction through linear interaction energy (LIE) theory. For ligand binding to flexible Cytochrome P450s (CYPs), this method was shown to decrease the root-mean-square error and standard deviation of error prediction by combining interaction energies of simulations starting from different conformations. Thereby, different parts of protein-ligand conformational space are sampled in parallel simulations. The iterative LIE framework relies on the assumption that separate simulations explore different local parts of phase space, and do not show transitions to other parts of configurational space that are already covered in parallel simulations. In this work, a method is proposed to (automatically) detect such transitions during the simulations that are performed to construct LIE models and to predict binding affinities. Using noise-canceling techniques and splines to fit time series of the raw data for the interaction energies, transitions during simulation between different parts of phase space are identified. Boolean selection criteria are then applied to determine which parts of the interaction energy trajectories are to be used as input for the LIE calculations. Here we show that this filtering approach benefits the predictive quality of our previous CYP 2D6-aryloxypropanolamine LIE model. In addition, an analysis is performed of the gain in computational efficiency that can be obtained from monitoring simulations using the proposed filtering method and by prematurely terminating simulations accordingly.     

C2-symmetric inhibitors of Plasmodium falciparum plasmepsin II: synthesis and theoretical predictions

A series of C(2)-symmetric compounds with a mannitol-based scaffold has been investigated, both theoretically and experimentally, as Plm II inhibitors. Four different stereoisomers with either benzyloxy or allyloxy P1/P1' side chains were studied. Computational ranking of the binding affinities of the eight compounds was carried out using the linear interaction energy (LIE) method relying on a complex previously determined by crystallography. Within both series of isomers the theoretical binding energies were in agreement with the enzymatic measurements, illustrating the power of the LIE method for the prediction of ligand affinities prior to synthesis. The structural models of the enzyme-inhibitor complexes obtained from the MD simulations provided a basis for interpretation of further structure-activity relationships. Hence, the affinity of a structurally similar ligand, but with a different P2/P2' substituent was examined using the same procedure. The predicted improvement in binding constant agreed well with experimental results.     

Comparison of end-point continuum-solvation methods for the calculation of protein-ligand binding free energies

We have compared the predictions of ligand‐binding affinities from several methods based on end‐point molecular dynamics simulations and continuum solvation, that is, methods related to MM/PBSA (molecular mechanics combined with Poisson–Boltzmann and surface area solvation). Two continuum‐solvation models were considered, viz., the Poisson–Boltzmann (PB) and generalised Born (GB) approaches. The nonelectrostatic energies were also obtained in two different ways, viz., either from the sum of the bonded, van der Waals, nonpolar solvation energies, and entropy terms (as in MM/PBSA), or from the scaled protein–ligand van der Waals interaction energy (as in the linear interaction energy approach, LIE). Three different approaches to calculate electrostatic energies were tested, viz., the sum of electrostatic interaction energies and polar solvation energies, obtained either from a single simulation of the complex or from three independent simulations of the complex, the free protein, and the free ligand, or the linear‐response approximation (LRA). Moreover, we investigated the effect of scaling the electrostatic interactions by an effective internal dielectric constant of the protein (?_int). All these methods were tested on the binding of seven biotin analogues to avidin and nine 3‐amidinobenzyl‐1H‐indole‐2‐carboxamide inhibitors to factor Xa. For avidin, the best results were obtained with a combination of the LIE nonelectrostatic energies with the MM+GB electrostatic energies from a single simulation, using ?_int = 4. For fXa, standard MM/GBSA, based on one simulation and using ?_int = 4–10 gave the best result. The optimum internal dielectric constant seems to be slightly higher with PB than with GB solvation. © Proteins 2012; © 2012 Wiley Periodicals, Inc.