Quantum mechanical, semiempirical (AM1) and ab initio (6-31G*) study of the Burkholderia cepacia lipase (BCL) catalysed reactions of the secondary alcohol esterification and its ester hydrolysis is presented. We have selected BCL for our study because of numerous experimental results available, but also because of its broad selectivity and stability that makes it interesting for industrial use. Previously we developed models for predicting lipase stereo-selectivity towards primary and secondary alcohols according to their structural parameters. In this work we show that not all of the experimentally determined binding modes are catalytically competent and that additional molecular modelling should be accomplished in order to find good starting points to study chemical reactions. The binding modes from which chemical modification of a substrate is possible are the most relevant for understanding enzyme selectivity and for the rational enzyme engineering.We also investigated the influence of the tetrahedral atom type, C and P, upon the energy barriers in the proton transfer reactions from the catalytic histidine (His286) to either the catalytic serine (Ser87) or the alcohol oxygen of the substrate.

The potential energy surface (PES) of 2-methylindole-3-acetic acid (2-Me-IAA) has been investigated via RHF/6-31G* and RHF/6-31++G** calculations. With both basis sets, only three symmetry unique local minima with syn orientation of the COOH group are contained in the PES. A network of reaction paths connects these energy minima. One symmetry unique energy minimum has an extremely low barrier to the adjacent global minimum. Comparison with unsubstituted IAA and its alkylated as well as halogenated derivatives substituted at positions 4, 5, 6, and 7 shows that the PES of 2-Me-IAA is quite different. This is in accord with its significantly lower auxin activity and weaker binding to auxin binding protein 1 (ABP1).

The activity of a biological compound is dependent both on specific binding to a target receptor and its ADME (Absorption, Distribution, Metabolism, Excretion) properties. A challenge to predict biological activity is to consider both contributions simultaneously in deriving quantitative models. We present a novel approach to derive QSAR models combining similarity analysis of molecular interaction fields (MIFs) with prediction of logP and/or logD. This new classification method is applied to a set of about 100 compounds related to the auxin plant hormone. The classification based on similarity of their interaction fields is more successful for the indole than the phenoxy compounds. The classification of the phenoxy compounds is however improved by taking into account the influence of the logP and/or the logD values on biological activity. With the new combined method, the majority (8 out of 10) of the previously misclassified derivatives of phenoxy acetic acid are classified in accord with their bioassays. The recently determined crystal structure of the auxin-binding protein 1 (ABP1) enabled validation of our approach. The results of docking a few auxin related compounds with different biological activity to ABP1 correlate well with the classification based on similarity of MIFs only. Biological activity is, however, better predicted by a combined similarity of MIFs + logP/logD approach.

The potential energy surface of 1-H-pyrrolo[3,2-b]pyridine-3-yl acetic acid has been investigated via RIIF/6-31G* calculations. The stationary points and reaction paths for syn orientation of the COOH group were determined and are compared with those of the derivatives of 3-indole acetic acid, which act as plant growth hormones. 1-H-pyrrolo[3,2-b]pyridine-3-yl acetic acid forms a kinetically stable conformer with a strong intramolecular hydrogen bond, in which the COOH group is in anti orientation. The influence of this hydrogen bond on bond lengths and vibration frequencies is described.

The conformational properties of N-formyl l-alanine amide (ALA) were investigated using RMP2/6-311G∗∗ ab initio gradient geometry optimization. One hundred forty four structures of ALA were optimized at 30° grid points in its φ(N–C(α)), ψ(C(α)–C′) conformational space. Using cubic spline functions, the grid structures were then used to construct analytical representations of complete surfaces, in φ,ψ-space, of bond lengths, bond angles, torsional sensitivity and electrostatic atomic charges. Analyses show that, in agreement with previous studies, the right-handed helical conformation, αR, is not a local energy minimum of the potential energy surface of ALA. Comparisons with protein crystallographic data show that the characteristic differences between geometrical trends in dipeptides and proteins, previously found for ab initio dipeptide structures obtained without electron correlation, are also found in the electron-correlated geometries. In contrast to generally accepted features of force fields used in empirical molecular modeling, partial atomic charges obtained by the CHELPG method are found to be not constant, but to vary significantly throughout the φ,ψ-space. By comparing RHF and MP2 structures, the effects of dispersion forces on ALA were studied, revealing molecular contractions for those conformations, in which small adjustments of torsional angles entail large changes in non-bonded distances.

Having measured the log P value of gamma -hydroxybutyric acid (GHB) in chloroform/water and dichloromethane/ water systems at T = 25 and 37 degreesC, the relative log P value for the structural analogue gamma -aminobutyric acid (GABA), log P (GABA) - log P (GHB), has been determined theoretically. A combined ab initio/Monte Carlo method for obtaining the internal free energy and the relative solvation free energy, respectively, has been applied for studying the conformational/tautomeric equilibria of GABA in the three solvents. Free energy changes for the partition processes have been calculated within thermodynamic cycles. There is no significant difference in the log P values for the neutral GABA and GHB of - 1.22 +/- 0.12 at 37 degreesC, as measured in the chloroform/water system. Using the more polar dichloromethane solvent as the organic phase, the GABA log P increases compared to its log P in chloroform. Because GABA almost entirely takes the zwitterionic form in aqueous solution, it is a key question whether this tautomeric form can partition into a slightly-polar phase. The lack of this ability would be in accord with the experimentally known fact that GABA does not cross the blood-brain barrier. Calculations indicated that the zwitterionic form is more stable than the neutral tautomer by as much as similar to9 kcal/mol in water at 37 degreesC. Direct partitioning of the zwitterion from the aqueous to the organic phase requires 23 or about 17 kcal/mol increases in free energy to enter the chloroform or dichloromethane phase, respectively. Although partitioning via the neutral form is allowed thermodynamically, equilibrium is reached after the entrance of about 3% of the neutral form into the organic phase. This partition leads to a negligible shift of the neutral form--zwitterion equilibrium in water, leaving essentially all GABA in the aqueous phase.

Since gamma-aminobutyric acid does not form a stable zwitterionic species in the gas phase, as calculated at the HF/6-311++G** level, the GABA.2H(2)O system was optimized for several neutral and zwitterionic GABA tantomers/conformers. The obtained molecular geometries and vibrational frequencies determined for the dihydrates reflect structural changes for GABA due to close and strongly bound water molecules. By use of GABA geometries optimized in the dihydrates, relative free energies of different species in aqueous solution were calculated. MP2/6-311++G**//HF/6-311++G"* energy values show that the neutral form is strongly preferred over the zwitterionic one for the isolated GABA species. The neutral tautomer, which is most stable in the gas phase, is only marginally changed by hydration; it is without an intramolecular hydrogen bond and has nearly gauche-gauche arrangements, 54 and -83 degrees, for the NCCC and the CCCC torsion angles, respectively, as determined in the dihydrate; In aqueous solution the zwitterionic structure is dominant. Comparison of cyclic gauche-gauche forms and a partially extended, gauche-trans structure indicates the preference of the more extended form. This structure differs from the trans-gauche zwitterionic conformer found for GABA by X-ray experiments in the crystalline phase. The experimental GABA conformer is not stable either in the isolated form or in the gas-phase dihydrate. It is, however, more stable by about 6.5 kcal/mol than the gas-phase gauche-trans form, as turned out in a restricted geometry optimization. Such a large internal stabilization may allow the existence (even preference) of the experimental zwitterionic GABA structure in aqueous solution if solvent effects are preferable. Partition of GABA between water and chloroform is not favored. At least 7.5 kcal/mol free energy increase is required if the zwitterion either directly or after transformation to a neutral form would leave the aqueous phase and enter chloroform. This result. supports the experimental finding that GABA does not cross the blood-brain barrier.

To study the nonplanarity of peptide bonds, the conformationally dependent variations of the N-C torsional angle, omega (2) (Figure 1), of the central peptide group in N-formyl L-alanyl L-alanine amide (ALA-ALA) was investigated using a database of Il 664 RHF/4-21G ab initio gradient optimized structures. The database was generated by optimizing the geometries of ALA-ALA at grid points in its four-dimensional (phi (1),psi (1),phi (2),psi (2)) conformational space (Figure I) defined by 40 degrees increments along the outer torsions phi (1) and psi (2), and by 30 degrees increments along the inner torsions psi (1) and phi (2). Using cubic spline functions, the grid structures were then used to construct analytical representations of complete surfaces of omega (2) in (phi (1),psi (1),phi (2),psi (2))-space. Analyses of the conformational surfaces of oz reveal that the peptide N-C torsion is a smoothly varying function of associated phi and psi angles and that, for many conformational regions, deviations from planarity are the rule rather than the exception. Comparisons with protein crystallographic data show that, in contrast to peptide torsional angles calculated for an entire protein, the omega (2) angles of smaller model peptides, such as ALA-ALA, cannot be used to model peptide groups in proteins, because of lone-range effects present in the latter but not the former. This finding indicates the general difficulty of predicting the exact positions of backbone torsional angles in proteins from smaller model peptides. Furthermore, the results confirm the directional nature of polypeptide chains. That is, conformation transmission effects from neighboring groups differ, depending on whether they are transmitted from right to left or from left to right in the peptide chain.

The molecular structures of n-hexane were determined by RHF/4-21G ab initio geometry optimization at 30° grid points in its three-dimensional τ1(C11–C8–C5–C1), τ2(C14–C11–C8–C5), τ3(C17–C14–C11–C8) conformational space. Of the resulting 12×12×12=1728 grid structures, 468 are symmetrically non-equivalent and were optimized constraining the torsions τ1, τ2, and τ3 to the respective grid points, while all other structural parameters were relaxed without any constraints. From the results, complete parameter surfaces were constructed using natural cubic spline functions, which make it possible to calculate parameter gradients, |∇P|=[(∂P/∂τi)2+(∂P/∂τj)2]1/2, where P is a C–C bond length or C–C–C angle. The parameter gradients provide an effective measure of the torsional sensitivity of the system and indicate that dynamic activities in one part of the molecule can significantly affect the density of states, and thus the contributions to vibrational entropy, in another part. This opens the possibility of dynamic entropic conformational steering in complex molecules; i.e. the generation of free energy contributions from dynamic effects of one part of a molecule on another. When the conformational trends in the calculated C–C bond lengths and C–C–C angles are compared with average parameters taken from some 900 crystallographic structures containing n-hexyl fragments or longer C–C bond sequences, some correlation between calculated and experimental trends in angles is found, in contrast to the bond lengths for which the two sets of data are in complete disagreement. The results confirm experiences often made in crystallography. That is, effects of temperature, crystal structure and packing, and molecular volume effects are manifested more clearly in bond lengths than bond angles which depend mainly on intramolecular properties. Frequency analyses of the τ1, τ2 and τ3 torsional angles in the crystal structures show conformational steering in the sense that, if τ1 is trans peri-planar (170°≤τ1≤180°; −180°≤τ1≤−170°), the values of τ2 and τ3 are clustered closely around the ideal gauche (±60°) and trans (±180°) positions. In contrast, when τ1 is in the region (50°≤τ1≤70°), there is a definite increase in the populations of τ2 and τ3 at −90 and −150°.

Based on a database of 11 664 RHF/4-21G ab initio gradient-optimized structures of N-formyl-L-alanyl-L-alanine amide (ALA-ALA), the local geometries and torsional sensitivity of this compound were analyzed to test the dipeptide approximation frequently used in peptide conformational analyses. This database was generated by optimizing the geometries of this compound at grid points in its four-dimensional (phi(1),psi(1),phi(2),psi(2)) conformational space defined by 40 degrees increments along the outer torsions phi(1) and and by 30 degrees increments along the inner torsions psi(1) and phi(2). Using cubic spline functions, the grid structures were then used to construct analytical representations of complete surfaces of the structural parameters of AL A-ALA, and of their gradients, in (phi(1),psi(1),phi(2),psi(2)) space. Analysis of the structural surfaces shows not only that the structure of a given residue in a peptide chain depends acutely on the conformational state of a neighboring residue but also that the interresidue effects differ, depending on whether they are transmitted from right to left or from left to right in the peptide chain. Structural gradients are a qualitative measure of the torsional sensitivity, and therefore of the density of states and contributions to vibrational entropy. Analyses of the gradient surfaces show that the density of states in a residue is significantly affected by the dynamics of a neighboring residue. This opens the possibility of dynamic entropic conformational steering in extended peptide chains, i.e., the generation of free energy contributions from dynamic effects of one part of a molecule on another, possibly stabilizing a conformational region of a PES whose static energy profile is less favorable compared to other regions. The gradient trends illustrate how the overall stability of a complex molecule is not only a function of how the static energy minima of its isolated subunits combine but also of how the dynamics of the subunits interact with one other. These interactions between individual residues represent a hidden cooperative effect that is not apparent at all in the dynamics of isolated dipeptide units.

The geometries of 28 compounds of type X–C1–C2–C3–Y, with X,Y=CH3, F, Cl, OH, NH2, COH, and COOH, were fully optimized by ab initio HF/4-21G calculations at 30° grid points in their respective φ(X–C1–C2–C3), ψ(C1–C2–C3–Y)-torsional spaces. The results make it possible to construct parameter surfaces and their gradients in φ,ψ-space. The magnitude of the gradient, |∇P|=[(∂P/∂φ)2+(∂P/∂ψ)2]1/2, of a structural parameter P (a bond length, bond angle, or non-bonded distance) in φ,ψ-torsional space is a measure of torsional sensitivity (TS); i.e. a measure of the extent to which bond lengths, bond angles, and non-bonded distances change at a point in φ,ψ-space with backbone torsional angles. It is found that TS is not constant throughout the conformational space of a molecule, but varies in a characteristic way. It seems that, regardless of the nature of X or Y, extended forms are typically in regions of low TS; puckered conformations, of high TS. Conformations with two sequential gauche torsional angles (GG sequences) are characterized by high TS of 1,5-non-bonded distances concomitant with relatively low TS of other internal coordinates. This property of GG sequences is the source of a stabilizing and cooperative energy increment that is not afforded by other torsional sequences, such as trans–trans or trans–gauche. A structural data base, consisting of thousands of HF/4-21G structures of X–C–C–Y and X–C–C–C–Y systems has been assembled and is available on a CD.

The potential energy surface of 4-methyl-3-indole acetic acid (IAA) was investigated via RHF/6-31G* calculations. Five symmetry-unique local minima with syn-periplanar orientation of the COOH group are present. In contrast, unsubstituted IAA and 4-Cl-IAA have only four such minima. The shape of the energy surface is different in all three cases, and that of 4-Me-IAA is found to contain elements of both reference compounds. Semiempirical energy surfaces (AM1, PM3, MNDO, and MINDO/3), which were determined too, differ significantly among each other and from the restricted Hartree-Fock (RHF) one. (C) 1999 John Wiley & Sons, Inc.

RHF/G-31G* investigations of 4-, 5-, and 6-ethyl(Et)-indole-3-acetic acid (IAA) yielded 11 symmetry-unique local minima with syn-periplanar orientation of the -COOH group for each of these compounds. The global minima are of C-1 symmetry in all cases. Comparison with earlier results shows that ethylation or chlorination in position 5 or 6 introduces only minor changes on the orientation of the acetic acid side group, with no effect on the reaction paths related to this group. For 4-Et-IAA, the deviations from unsubstituted IAA are larger but preserve the pattern of reaction paths that is present in unsubstituted IAA,which is in contrast to 4-Cl-IAA, where local minima and reaction paths are completely different. (C) 1998 John Wiley & Sons, Inc.

Together with ab initio calculations, Raman spectra of the methylated disilanes MeX2SiSiX2Me (X=H, F, Br and I) unambiguously prove the presence of two rotamers, anti and gauche in the liquid state. Energy differences between rotational isomers have been determined from variable temperature Raman spectra employing Van't Hoff plots of the logarithm of intensity ratios vs. the inverse temperature. For the disilanes with X=F, Br and I the anti conformation is preferred in the liquid state. For MeH2SiSiH2Me the gauche rotamer is slightly stabilized by 0.43 kJ mol−1. Vibrational spectra have been assigned by the help of normal coordinate analyses using optimized geometries and ab initio symmetry force fields.

Detailed conformational analyses of propionic and isobutyric acids were performed to contribute to a better understanding of the stereochemical characteristics of biologically active indole-3-aliphatic acids. The studies are based on ab initio SCF (RHF/6-31G*) and molecular mechanics (force fields used: MM2, MM3, CFFS1, AMBER, CVFF, ESFF) methods. The results obtained with the CFFS1 and MM3 force fields revealed the best agreement with the experimental values and those from ab initio calculations. Normal mode frequencies in the harmonic oscillator approximation were calculated for the geometry optimized conformers with C-s symmetry of these compounds as well as of indole-3-acetic acid (IAA) and some of its biologically important derivatives (4-Cl-IAA, 6-Cl-IAA, 7-Cl-IAA, 4-Me-IAA) and indole-3-isobutyric acid (IIBA). The influence of the indole ring on the C=O and O-H stretching frequencies was analyzed. A small decrease of the C=O frequency was determined in the indole-3-acetic acid derivatives and a larger one in indole-3-isobutyric acid.

The ab initio conformational energy minima of the model tripeptide N-formyl-L-alanyl-L-alanine amide (ALA-ALA) were determined by ab initio RHF/4-21G and RHF/6-31G* gradient geometry refinement. For the current investigation, 11 664 RHF/4-21G structures were optimized, representing grid points in the four-dimensional (φ1, ψ1, φ2, ψ2) conformational space, which were constructed in 40∞ increments along the outer torsions φ1 and ψ2 and in 30∞ increments along the inner torsions ψ1 and φ2 of ALA-ALA. Two new energy minima, previously not reported, are described. The positions of the RHF/6-31G* energy minima in φ,ψ-space can differ significantly from the corresponding RHF/4-21G locations, and both sets are not clustered in the centers but on the fringes of the most populated regions of φ,ψ-space in protein crystal structures. Thus, the torsion angles of the ab initio energy minima are not those of the typical substructures of proteins: the most stable helices are not αR, and the torsion angles of the most stable bend forms are not those most frequently encountered in protein bends. Limitations of the dipeptide approximation are explored, illustrating how the conformational energies of an amino acid residue depend on the state of its neighbor.

Local minima and transition states were determined for 5-Cl-IAA, 6-Cl-IAA, 7-Cl-IAA and 5,6-Cl2-IAA via RHF/6-31G* optimizations. The results of these chlorinated IAAs are very similar to the results obtained for indole-3-acetic acid (IAA), but significantly differ from those obtained earlier for 4-Cl-IAA. The results are compared and discussed in view of the auxin activity, which is quite diverse in this set of compounds. In combination with the experimental data obtained by others, they enable us to make some assumptions regarding the biological activity. It seems that the auxin activity of the compounds in the present set is more discriminating by their ability to bind to the receptor, determined primarily by the position and nature of the subtituents than by transport or stability of these molecules in plants.

The energetically low-lying parts of the potential energy surface of L-proline were investigated by ab initio (RHF/6-311++G**)calculations. The results are discussed with respect to the parametrization of the MM3 force field and in comparison with those obtained earlier for glycine and alpha-alanine, on the one hand, and for N-acetyl-L-proline amide, on the other hand. (C) 1997 John Wiley & Sons, Inc.

Equilibrial composition of eight conformers of gamma-hydroxybutyric acid have been established in the gas phase and in aqueous solution by combining ab initio calculations and Monte Carlo simulations. Ab initio RHF/6-311++G** and MP2/6-311++G** optimized conformers have different relative free energies in the gas phase. On the basis of the MP2 results, there are at least four conformers in the gas-phase mixture. The most stable structure is a cyclic conformer without intramolecular hydrogen bond. The equilibrium composition in the aqueous solution is dominated by structures with extended conformations. The conformational preference is the result of a subtle balance of internal free energy and solvation terms. The extended structure is the favored one in a mixed solvent comprised of methanol and chloroform, as well. Calculations predict 1.9 and 2.3 kcal/mol activation free energy for the formation of a cyclic conformer, prerequisite for lactonization, in aqueous solution and in the methanol/chloroform solvent, respectively.

The conformational geometry changes associated with torsional motion about C-C single bonds are reported for some 47 basic organic molecules of the type (X,Y,Z)C-C(X',Y',Z'), and for the phi,psi space of the model dipeptides N-acetyl N'-methyl amides of glycine and alanine. All structures were determined by ab initio HF/4-21G geometry optimizations, Employing natural cubic spline parameters, a program was written for use in empirical modeling parameter development, which calculates the bond lengths and angles of the reported compounds as functions of the associated torsional angles at any point in conformational space. The additivity of the conformational geometry functions is explored and illustrates how cooperative effects emerge in complex molecules. The results are instructive for procedures in which the properties of molecular fragments are used to derive force field parameters for the empirical modeling of complex molecules, (C) 1997 Elsevier Science B.V.