Strong and weak hydrogen bonds in protein-ligand complexes of kinases: a comparative study

Strong and weak hydrogen bonds between protein and ligand are analyzed in a group of 233 X-ray crystal structures of the kinase family. These kinases are from both eukaryotic and prokaryotic organisms. The dataset comprises of 44 sub-families, out of which 35 are of human origin and the rest belong to other organisms. Interaction analysis was carried out in the active sites, defined here as a sphere of 10 A radius around the ligand. A majority of the interactions are observed between the main chain of the protein and the ligand atoms. As a donor, the ligand frequently interacts with amino acid residues like Leu, Glu and His. As an acceptor, the ligand interacts often with Gly, and Leu. Strong hydrogen bonds N-H...O, O-H...O, N-H...N and weak bonds C-H...O, C-H...N are common between the protein and ligand. The hydrogen bond donor capacity of Gly in N-H...O and C-H...O interactions is noteworthy. Similarly, the acceptor capacity of main chain Glu is ubiquitous in several kinase sub-families. Hydrogen bonds between protein and ligand form characteristic hydrogen bond patterns (supramolecular synthons). These synthon patterns are unique to each sub-family. The synthon locations are conserved across sub-families due to a higher percentage of conserved sequences in the active sites. The nature of active site water molecules was studied through a novel classification scheme, based on the extent of exposure of water molecules. Water which is least exposed usually participates in hydrogen bond formation with the ligand. These findings will help structural biologists, crystallographers and medicinal chemists to design better kinase inhibitors.     

Thermodynamic consequences of disrupting a water-mediated hydrogen bond network in a protein:pheromone complex

The mouse pheromones (+/-)-2-sec-butyl-4,5-dihydrothiazole (SBT) and 6-hydroxy-6-methyl-3-heptanone (HMH) bind into an occluded hydrophobic cavity in the mouse major urinary protein (MUP-1). Although the ligands are structurally unrelated, in both cases binding is accompanied by formation of a similar buried, water-mediated hydrogen bond network between the ligand and several backbone and side chain groups on the protein. To investigate the energetic contribution of this hydrogen bond network to ligand binding, we have applied isothermal titration calorimetry to measure the binding thermodynamics using several MUP mutants and ligand analogs. Mutation of Tyr-120 to Phe, which disrupts a hydrogen bond from the phenolic hydroxyl group of Tyr-120 to one of the bound water molecules, results in a substantial loss of favorable binding enthalpy, which is partially compensated by a favorable change in binding entropy. A similar thermodynamic effect was observed when the hydrogen bonded nitrogen atom of the heterocyclic ligand was replaced by a methyne group. Several other modifications of the protein or ligand had smaller effects on the binding thermodynamics. The data provide supporting evidence for the role of the hydrogen bond network in stabilizing the complex.     

Trans-substitution of the proximal hydrogen bond in myoglobin: II. Energetics, functional consequences, and implications for hemoglobin allostery

The trans-substituted histidine to glycine mutant of sperm whale myoglobin (H93G Mb) is used to study energetics of proximal hydrogen bonding, proximal ligand-heme interactions, and coupling to distal ligand binding. Comparison of mono- and dimethylimidazole structural isomers shows that the hydrogen bond between the proximal ligand and the neighboring Ser92 hydroxyl (position F7) is stabilizing. The range of hydrogen bond stabilities measured here for different distal ligand complexes ranges from -0.7 kcal/mol (monomethylimidazole isomers to MbCO) to -4.1 kcal/mol (dimethylimidazole isomers to MbCN). This range of hydrogen bond stabilities, which is similar to that seen in protein mutagenesis unfolding studies, demonstrates the high sensitivity of the hydrogen bond to modest structural perturbations. The degree to which the 2-methyl group destabilizes proximal ligand binding is found to depend inversely on the total electronic spin. For monomethylimidazole proximal ligands, distal ligand binding weakens the proximal hydrogen bond compared to deoxyMb. Surprisingly, this trend is largely reversed for the dimethylimidazole proximal ligands. These results demonstrate strong coupling between the proximal protein matrix and distal ligand binding. These results provide an explanation for the strong avoidance of hydrogen bonding residues at position F7 in hemoglobin sequences.     

N-H...O, O-H...O, and C-H...O hydrogen bonds in protein-ligand complexes: strong and weak interactions in molecular recognition

The characteristics of N-H...O, O-H...O, and C-H...O hydrogen bonds are examined in a group of 28 high-resolution crystal structures of protein-ligand complexes from the Protein Data Bank and compared with interactions found in small-molecule crystal structures from the Cambridge Structural Database. It is found that both strong and weak hydrogen bonds are involved in ligand binding. Because of the prevalence of multifurcation, the restrictive geometrical criteria set up for hydrogen bonds in small-molecule crystal structures may need to be relaxed in macromolecular structures. For example, there are definite deviations from linearity for the hydrogen bonds in protein-ligand complexes. The formation of C-H...O hydrogen bonds is influenced by the activation of the C(alpha)-H atoms and by the flexibility of the side-chain atoms. In contrast to small-molecule structures, anticooperative geometries are common in the macromolecular structures studied here, and there is a gradual lengthening as the extent of furcation increases. C-H...O bonds formed by Gly, Phe, and Tyr residues are noteworthy. The numbers of hydrogen bond donors and acceptors agree with Lipinski's "rule of five" that predicts drug-like properties. Hydrogen bonds formed by water are also seen to be relevant in ligand binding. Ligand C-H...O(w) interactions are abundant when compared to N-H...O(w) and O-H...O(w). This suggests that ligands prefer to use their stronger hydrogen bond capabilities for use with the protein residues, leaving the weaker interactions to bind with water. In summary, the interplay between strong and weak interactions in ligand binding possibly leads to a satisfactory enthalpy-entropy balance. The implications of these results to crystallographic refinement and molecular dynamics software are discussed.     

Investigation of the PDZ domain ligand binding site using chemically modified peptides

Several chemically modified analogues to a tightly binding ligand for the second PDZ domain of MAGI-3 were synthesized and evaluated for their ability to compete with native peptide ligands. N-methyl scanning of the ligand backbone amides revealed the energetically important hydrogen bonds between the ligand backbone and the PDZ domain. Analogues to the ligand's conserved threonine/serine(-2) residue, involved in a side chain to side chain hydrogen bond with a conserved histidine in the PDZ domain, revealed that the interaction is highly sensitive to the steric structure around the hydroxyl group of this residue. Analogues of the ligand carboxy terminus revealed that the full hydrogen bond network of the GLGF loop is important in ligand binding.     

HBAT: a complete package for analysing strong and weak hydrogen bonds in macromolecular crystal structures

The program HBAT is a tool to automate the analysis of potential hydrogen bonds and similar type of weak interactions like halogen bonds and non-canonical interactions in macromolecular structures, available in Brookhaven Protein Database (PDB) file format. HBAT is written using PERL and TK languages. The program generates an MSOFFICE Excel compatible output file for statistical analysis. HBAT identify potential interactions based on geometrical criteria. A series of analysis reports like frequency tables, geometry distribution tables, furcations list are generated. A user friendly GUI offers freedom to select several parameters and options. Graphviz based visualization of hydrogen bond networks in 2D helps to study the cooperativity and anticooperativity geometry in hydrogen bond. HBAT supports post docking interaction analysis between PDB files for any target/receptor (in PDB files) and docked ligands/poses (in SDF). This tool can be implemented in active site interaction analysis, structure based drug design and molecular dynamics simulations.     

Trans-substitution of the proximal hydrogen bond in myoglobin: I. Structural consequences of hydrogen bond deletion

The structural role of a side-chain to side-chain protein hydrogen bond is examined using trans-substitution of the proximal histidine of myoglobin with methylimidazoles (Barrick, Biochemistry 1994;33:6546-6554). Modification of the chemical structure of exogenous ligands allows this hydrogen bond to be disrupted. Comparison of the crystal structures of H93G myoglobin complexed 4-methylimidazole (4meimd; methylation at carbon 4) and 1-methylimidazole (1meimd; methylation at the adjacent nitrogen, preventing hydrogen bonding between the imidazole ligand and the protein) shows that the polypeptide, heme, and methylimidazole orientations are the same within error. For 4meimd there appear to be major and minor conformations corresponding to different tautomeric states of the ligand. Conformational heterogeneity is also seen in the hyperfine-shifted region of the NMR spectrum of 4meimd complexed with high-spin H93G deoxyMb. The major conformation of the 4meimd ligand and the 1meimd ligand, as seen in the respective crystal structures, are quite similar except that the proximal ligand NH-to-Ser92-OH hydrogen bond is eliminated in the 1meimd complex, and instead the proximal ligand CH is adjacent to the Ser92-OH. Thus, this system provides a means to eliminate the Mb proximal hydrogen bond in a chemically and structurally conservative way.     

A hydrogen bond regulates slow motions in ubiquitin by modulating a β-turn flip

Proteins exist as conformational ensembles composed of multiple interchanging substates separated by kinetic barriers. Interconverting conformations are often difficult to probe, owing to their sparse population and transient nature. Here, we report the identification and characterization of a subset of conformations in ubiquitin that participate in microsecond-to-millisecond motions in the amides of Ile23, Asn25, and Thr55. A novel side chain to the backbone hydrogen bond that regulates these motions has also been identified. Combining our NMR studies with the available X-ray data, we have unearthed the physical process underlying slow motions—the interconversion of a type I into a type II β-turn flip at residues Glu51 through Arg54. Interestingly, the dominant conformer of wild-type ubiquitin observed in solution near neutral pH is only represented by about 22% of the crystal structures. The conformers generated as a result of the dynamics of the hydrogen bond appear to be correlated to ligand recognition by ubiquitin.     

Ligand-based and structure-based approaches in identifying ideal pharmacophore against c-Jun N-terminal kinase-3

Structure and ligand based pharmacophore modeling and docking studies carried out using diversified set of c-Jun N-terminal kinase-3 (JNK3) inhibitors are presented in this paper. Ligand based pharmacophore model (LBPM) was developed for 106 inhibitors of JNK3 using a training set of 21 compounds to reveal structural and chemical features necessary for these molecules to inhibit JNK3. Hypo1 consisted of two hydrogen bond acceptors (HBA), one hydrogen bond donor (HBD), and a hydrophobic (HY) feature with a correlation coefficient (r²) of 0.950. This pharmacophore model was validated using test set containing 85 inhibitors and had a good r² of 0.846. All the molecules were docked using Glide software and interestingly, all the docked conformations showed hydrogen bond interactions with important hinge region amino acids (Gln155 and Met149) and these interactions were compared with Hypo1 features. The results of ligand based pharmacophore model (LBPM) and docking studies are validated each other. The structure based pharmacophore model (SBPM) studies have identified additional features, two hydrogen bond donors and one hydrogen bond acceptor. The combination of these methodologies is useful in designing ideal pharmacophore which provides a powerful tool for the discovery of novel and selective JNK3 inhibitors.     

1.25 A resolution crystal structures of human haemoglobin in the oxy, deoxy and carbonmonoxy forms

The most recent refinement of the crystallographic structure of oxyhaemoglobin (oxyHb) was completed in 1983, and differences between this real-space refined model and later R state models have been interpreted as evidence of crystallisation artefacts, or numerous sub-states. We have refined models of deoxy, oxy and carbonmonoxy Hb to 1.25 A resolution each, and compare them with other Hb structures. It is shown that the older structures reflect the software used in refinement, and many differences with newer structures are unlikely to be physiologically relevant. The improved accuracy of our models clarifies the disagreement between NMR and X-ray studies of oxyHb, the NMR experiments suggesting a hydrogen bond to exist between the distal histidine and oxygen ligand of both the alpha and beta-subunits. The high-resolution crystal structure also reveals a hydrogen bond in both subunit types, but with subtly different geometry which may explain the very different behaviour when this residue is mutated to glycine in alpha or beta globin. We also propose a new set of relatively fixed residues to act as a frame of reference; this set contains a similar number of atoms to the well-known "BGH" frame yet shows a much smaller rmsd value between R and T state models of HbA.     

Ligand-induced strain in hydrogen bonds of the c-Src SH3 domain detected by NMR

Changes in the molecular conformation of proteins can result from a variety of perturbations, and can play crucial roles in the regulation of biological activity. A new solution NMR method has been applied to monitor ligand-induced changes in hydrogen bond geometry in the chicken c-Src SH3 domain. The structural response of this domain to ligand binding has been investigated by measuring trans-hydrogen bond (15)N-(13)C' scalar couplings in the free state and when bound to the high affinity class I ligand RLP2, containing residues RALPPLPRY. A comparison between hydrogen bonds in high resolution X-ray structures of this domain and those observed via (h3)J(NC') couplings in solution shows remarkable agreement. Two backbone-to-side-chain hydrogen bonds are observed in solution, and each appears to play a role in stabilization of loop structure. Reproducible ligand-induced changes in trans-hydrogen bond scalar couplings are observed across the domain that translate into changes in hydrogen bond length ranging between 0.02 to 0.12 A. The observed changes can be rationalized by an induced fit mechanism in which hydrogen bonds across the protein participate in a compensatory response to forces imparted at the protein-ligand interface. Upon ligand binding, mutual intercalation of the two Leu-Pro segments of the ligand between three aromatic side-chains protruding from the SH3 surface wedges apart secondary structural elements within the SH3 domain. This disruption is transmitted in a domino-like effect across the domain through networks of hydrogen bonded peptide planes. The unprecedented resolution obtained demonstrates the ability to characterize subtle structural rearrangements within a protein upon perturbation, and represents a new step in the endeavor to understand how hydrogen bonds contribute to the stabilization and function of biological macromolecules.     

The MOSIX Direct File System Access Method for Supporting Scalable Cluster File Systems

MOSIX is a cluster management system that supports preemptive process migration. This paper presents the MOSIX Direct File System Access (DFSA), a provision that can improve the performance of cluster file systems by allowing a migrated process to directly access files in its current location. This capability, when combined with an appropriate file system, could substantially increase the I/O performance and reduce the network congestion by migrating an I/O intensive process to a file server rather than the traditional way of bringing the file's data to the process. DFSA is suitable for clusters that manage a pool of shared disks among multiple machines. With DFSA, it is possible to migrate parallel processes from a client node to file servers for parallel access to different files. Any consistent file system can be adjusted to work with DFSA. To test its performance, we developed the MOSIX File-System (MFS) which allows consistent parallel operations on different files. The paper describes DFSA and presents the performance of MFS with and without DFSA.     

Strong and weak hydrogen bonds in drug-DNA complexes: A statistical analysis

A statistical analysis of strong and weak hydrogen bonds in the minor groove of DNA was carried out for a set of 70 drug-DNA complexes. The terms ‘strong’ and ‘weak’ pertain to the inherent strengths and weakness of the donor and acceptor fragments rather than to any energy considerations. The dataset was extracted from the protein data bank (PDB). The analysis was performed with an in-house software, hydrogen bond analysis tool (HBAT). In addition to strong hydrogen bonds such as O—H⋯O and N—H⋯O, the ubiquitous presence of weak hydrogen bonds such as C—H⋯O is implicated in molecular recognition. On an average, there are 1.4 weak hydrogen bonds for every strong hydrogen bond. For both categories of interaction, the N(3) of purine and the O(2) of pyrimidine are favoured acceptors. Donor multifurcation is common with the donors generally present in the drug molecules, and shared by hydrogen bond acceptors in the minor groove. Bifurcation and trifurcation are most commonly observed. The metrics for strong hydrogen bonds are consistent with established trends. The geometries are variable for weak hydrogen bonds. A database of recognition geometries for 26 literature amidinium-based inhibitors of Human African Trypanosomes (HAT) was generated with a docking study using seven inhibitors which occur in published crystal structures included in the list of 70 complexes mentioned above, and 19 inhibitors for which the drug-DNA complex crystal structures are unknown. The virtual geometries so generated correlate well with published activities for these 26 inhibitors, justifying our assumption that strong and weak hydrogen bonds are optimized in the active site.     

The crystal structures of Phascolopsis gouldii wild type and L98Y methemerythrins: structural and functional alterations of the O2 binding pocket

Reported are the X-ray crystal structures of recombinant Phascolopsis gouldii methemerythrin (1.8-A resolution) and the structure of an O2-binding-pocket mutant, L98Y methemerythrin (2.1-A resolution). The L98Y hemerythrin (Hr) has a greatly enhanced O2 affinity, a slower O2 dissociation rate, a larger solvent deuterium isotope effect on this rate, and a greater resistance to autoxidation relative to the wild-type protein. The crystal structures show that the hydrophobic binding pocket of Hr can accommodate substitution of a leucyl by a tyrosyl side chain with relatively minor structural rearrangements. UV/vis and resonance Raman spectra show that in solution L98Y methemerythrin contains a mixture of two diiron site structures differing by the absence or presence of an Fe(III)-coordinated phenolate. However, in the crystal, only one L98Y diiron site structure is seen, in which the Y98 hydroxyl is not a ligand, but instead forms a hydrogen bond to a terminal hydroxo/aqua ligand to the nearest iron. Based on this crystal structure, we propose that in the oxy form of L98Y hemerythrin the non-polar nature of the binding pocket favors localization of the Y98 hydroxyl near the O2 binding site, where it can donate a hydrogen bond to the hydroperoxo ligand. The stabilizing Y98OH-O2H-interaction would account for all of the altered O2 binding properties of L98Y Hr listed above.     

Thermodynamics of hydrogen bond and hydrophobic interactions in cyclodextrin complexes.

Values of K, delta G(o), delta H(o), delta S(o) and delta C(po) for the binding reaction of small organic ligands forming 1:1 complexes with either alpha- or beta-cyclodextrin were obtained by titration calorimetry from 15 degrees C to 45 degrees C. A hydrogen bond or hydrophobic interaction was introduced by adding a single functional group to the ligand. The thermodynamics of binding with and without the added group are compared to estimate the contribution of the hydrogen bond or hydrophobic interaction. A change in the environment of a functional group is required to influence the binding thermodynamics, but molecular size-dependent solute-solvent interactions have no effect. For phenolic O-H-O hydrogen bond formation, delta H(o) varies from -2 to -1.4 kcal mol(-1) from 15 degrees C to 45 degrees C, and delta C(p) is increased by 18 cal K(-1) mol(-1). The hydrophobic interaction has an opposite effect: in alpha-cyclodextrin, delta C(po) = -13.3 cal K(-1) mol(-1) per ligand -CH(2)-, identical to values found for the transfer of a -CH(2)-group from water to a nonpolar environment. At room temperature, the hydrogen bond and the -CH(2)-interaction each contribute about -600 cal mol(-1) to the stability (delta G(o)) of the complex. With increased temperature, the hydrogen bond stability decreases (i.e., hydrogen bonds "melt"), but the stability of the hydrophobic interaction remains essentially constant.     

Differential ligand recognition by the Src and phosphatidylinositol 3-kinase Src homology 3 domains: circular dichroism and ultraviolet resonance Raman studies

Src homology 3 (SH3) domains recognize Pro-rich motifs using a hydrophobic cleft which contains several conserved aromatic residues. To investigate how aromatic residues contribute to ligand recognition, circular dichroism (CD) and 235 nm excited ultraviolet resonance Raman spectroscopies have been applied to Src and phosphatidylinositol 3-kinase (PI3K) SH3s. The CD analysis shows that Src SH3 binds to RPLPPLP (R-core) using aromatic residues with a dissociation constant (K(d)) of 10 microM. Moreover, intensity increases of the Trp and Tyr Raman bands suggest that the interaction is mediated by hydrophobic contacts and/or hydrogen bond formation with both Trp and Tyr residues. In the interaction of Src SH3 with VSLARRPLPPLP (VSL12) (K(d) 0.8 microM), Trp118 appears to form a strong hydrogen bond with VSL12, judging from significant intensity increases of the Trp Raman bands and the reported complex structure. In contrast, PI3K SH3 binds to R-core and VSL12 with lower affinities (K(d) 34 and 18 microM, respectively), and the interactions are suggested to be mediated mainly by hydrophobic contacts and/or hydrogen bond formation with Tyr residue(s). In the D21N mutant (Asp21 --> Asn) of PI3K SH3, whose hydrophobic cleft is deformed, Trp55 is shown to be responsible for the interaction with VSL12 by intensity increases of the Trp Raman bands. However, the affinity is severely decreased (K(d) 330 microM). These observations imply that SH3 domains associate with their ligands with distinct use of aromatic residues and that hydrogen bond formation with an SH3-conserved Trp residue in the well-ordered hydrophobic cleft is important for stable complex formation.     

The Specific Cleavage of Lactone Linkage to Open-Loop in Cyclic Lipopeptide during Negative ESI Tandem Mass Spectrometry: The Hydrogen Bond Interaction Effect of 4-Ethyl Guaiacol

Mass spectrometry is a valuable tool for the analysis and identification of chemical compounds, particularly proteins and peptides. Lichenysins G, the major cyclic lipopeptide of lichenysin, and the non-covalent complex of lichenysins G and 4-ethylguaiacol were investigated with negative ion ESI tandem mass spectrometry. The different fragmentation mechanisms for these compounds were investigated. Our study shows the 4-ethylguaiacol hydrogen bond with the carbonyl oxygen of the ester group in the loop of lichenysins G. With the help of this hydrogen bond interaction, the ring structure preferentially opens in lactone linkage rather than O-C bond of the ester-group to produce alcohol and ketene. Isothermal titration ¹H-NMR analysis verified the hydrogen bond and determined the proportion of subject and ligand in the non-covalent complex to be 1∶1. Theoretical calculations also suggest that the addition of the ligand can affect the energy of the transition structures (TS) during loop opening.     

The role of hydrogen bonding at the active site of a cupredoxin: the Phe114Pro azurin variant

The Phe114Pro mutation to the cupredoxin azurin (AZ) leads to a number of structural changes at the active site attributed to deletion of one of the hydrogen bonds to the Cys112 ligand, removal of the bulky phenyl group from the hydrophobic patch of the protein, and steric interactions made by the introduced Pro. The remaining hydrogen bond between the coordinating thiolate and the backbone amide of Asn47 is strengthened. At the type-1 copper site, the Cu(II)-O(Gly45) axial interaction decreases, while the metal moves out of the plane formed by the equatorial His46, Cys112, and His117 ligands, shortening the bond to the axially coordinating Met121. The resulting distorted tetrahedral geometry is distinct from the trigonal bipyramidal arrangement in the wild-type (WT) protein. The unique position of the main S(Cys) --> Cu(II) ligand-to-metal charge-transfer transition in AZ (628 nm) has shifted in the Phe114Pro variant to a value that is more typical for cupredoxins (599 nm). This probably occurs because of the removal of the Phe114-Cys112 hydrogen bond. The Phe114Pro mutation results in a 90 mV decrease in the reduction potential of AZ, and removal of the second hydrogen bond to the Cys ligand seems to be the major cause of this change. The C-terminal His117 ligand does not protonate in the reduced Phe114Pro AZ variant, which suggests that none of the structural features altered by the mutation are responsible for the absence of this effect in the WT protein. Upon reduction, the copper displaces further from the equatorial ligand plane and the Cu-S(Met121) bond length decreases. These changes are larger than those seen in the WT protein and contribute to the order of magnitude decrease in the intrinsic electron-transfer capabilities of the Phe114Pro variant.     

Six-coordinate phosphorus compounds. Crystal structures and dynamic NMR studies of phosphoranes containing chelating ligands

Hexacoordination of the neutral phosphorus compounds 4–6 is evidenced by their high field ³¹P NMR chemical shifts and is further substantiated by the crystal structure of 5 and 6.5 contains the potentially bis-chelating ligand Ar = (C₆H₃(CH₂NMe₂)₂-2,6) and 6 the same ligand with a protonated amino group. In both cases the compounds exhibit slightly distorted octahedral geometry. In compound 5, only one NMe₂ group is coordinated to the phosphorus atom with an N → P bond of 2.063 Å. In compound 6, the NMe₂ group is coordinated to the phosphorus atom with an N → P bond of 2.007 Å while the dimethylammonium substituent is pointing away from the phosphorus atom forming a hydrogen bridge with two oxygen atoms. The fluxional behavior of these three novel six-coordinate phosphorus compounds was studied by dynamic ¹H NMR spectroscopy.