Christopher J. R. Illingworth

Assessing the Role of Polarization in Docking

We describe a strategy for including ligand and protein polarization in docking that is based on the conversion of induced dipoles to induced charges. Induced charges have a distinct advantage in that they are readily implemented into a number of different computer programs, including many docking programs and hybrid QM/MM programs; induced charges are also more readily interpreted. In this study, the ligand was treated quantum mechanically to avoid parametrization issues and was polarized by the target protein, which was treated as a set of point charges. The induced dipole at a given target atom, due to polarization by the ligand and neighboring residues, was reformulated as induced charges at the given atom and its bonded neighbors, and these were allowed to repolarize the ligand in an iterative manner. The final set of polarized charges was evaluated in docking using AutoDock 4.0 on 12 protein-ligand systems against the default empirical Gasteiger charges, and against nonpolarized and partially polarized potential-derived charges. One advantage of AutoDock is that the best rmsd structure can be identified not only from the lowest energy pose but also from the largest cluster of poses. Inclusion of polarization does not always lead to the lowest energy pose having a lower rmsd, because docking is designed by necessity to be rapid rather than accurate. However, whenever an improvement in methodology, corresponding to a more thorough treatment of polarization, resulted in an increased cluster size, then there was also a corresponding decrease in the rmsd. The options for implementing polarization within a purely classical docking framework are discussed.

Closed loop folding units from structural alignments: Experimental foldons revisited

Nonoverlapping closed loops of around 25–35 amino acids formed via nonlocal interactions at the loop ends have been proposed as an important unit of protein structure. This hypothesis is significant as such short loops can fold quickly and so would not be bound by the Leventhal paradox, giving insight into the possible nature of the funnel in protein folding. Previously, these closed loops have been identified either by sequence analysis (conservation and autocorrelation) or studies of the geometry of individual proteins. Given the potential significance of the closed loop hypothesis, we have explored a new strategy for determining closed loops from the insertions identified by the structural alignment of proteins sharing the same overall fold. We determined the locations of the closed loops in 37 pairs of proteins and obtained excellent agreement with previously published closed loops. The relevance of NMR structures to closed loop determination is briefly discussed. For cytochrome c, cytochrome b562 and triosephophate isomerase, independent folding units have been determined on the basis of hydrogen exchange experiments and misincorporation proton‐alkyl exchange experiments. The correspondence between these experimentally derived foldons and the theoretically derived closed loops indicates that the closed loop hypothesis may provide a useful framework for analyzing such experimental data.

Toward a consistent treatment of polarization in model QM/MM calculations.

The concept of model chemistries within hybrid QM/MM calculations has been addressed through analysis of the polarization energy determined by two distinct approaches based on (i) induced charges and (ii) induced dipoles. The quantum mechanical polarization energy for four configurations of the water dimer has been determined for a range of basis sets using Morokuma energy decomposition analysis. This benchmark value has been compared to the fully classical polarization energy determined using the induced dipole approach, and the molecular mechanics polarization energy calculated using induced charges within the MM region of hybrid QM/MM calculations. From the water dimer calculations, it is concluded that the induced charge approach is consistent with medium sized basis set calculations whereas the induced dipole approach is consistent with large basis set calculations. This result is highly relevant to the concept of QM/MM model chemistries.

Connectivity and binding‐site recognition: Applications relevant to drug design

Here, we describe a family of methods based on residue–residue connectivity for characterizing binding sites and apply variants of the method to various types of protein–ligand complexes including proteases, allosteric‐binding sites, correctly and incorrectly docked poses, and inhibitors of protein–protein interactions. Residues within ligand‐binding sites have about 25% more contact neighbors than surface residues in general; high‐connectivity residues are found in contact with the ligand in 84% of all complexes studied. In addition, a k‐means algorithm was developed that may be useful for identifying potential binding sites with no obvious geometric or connectivity features. The analysis was primarily carried out on 61 protein–ligand structures from the MEROPS protease database, 250 protein–ligand structures from the PDBSelect (25%), and 30 protein–protein complexes. Analysis of four proteases with crystal structures for multiple bound ligands has shown that residues with high connectivity tend to have less variable side‐chain conformation. The relevance to drug design is discussed in terms of identifying allosteric‐binding sites, distinguishing between alternative docked poses and designing protein interface inhibitors. Taken together, this data indicate that residue–residue connectivity is highly relevant to medicinal chemistry.

The effect of MM polarization on the QM/MM transition state stabilization: application to chorismate mutase

Hybrid quantum mechanics/molecular mechanics (QM/MM) calculations provide a mechanism for studying enzyme catalysed reactions at the molecular level. Here, through applications on the chorismate to prephenate rearrangement within the enzyme chorismate mutase, the feasibility of including MM polarization into these calculations has been demonstrated using the method of induced charges. MM polarization is shown to be a short-range effect, such that more than 75% of the energy of MM polarization occurs within a 5 Å residue-based cut-off of the substrate. MM polarization was shown to have a greater magnitude within the enzyme catalysed reaction than in the aqueous reaction, indicating that MM polarization may in principle have a significant effect on enzyme rate enhancement and mechanism. In both the enzyme and the aqueous case, the percentage contribution of MM polarization to the total stabilization energy was towards the upper end of the expected value. For the specific structures studied here, MM polarization lowered the energy barrier for the aqueous reaction, but the calculated contribution of MM polarization to both the reactant and transition structure stability were similar.

Assessing the effect of dynamics on the closed-loop protein-folding hypothesis

The closed-loop (loop-n-lock) hypothesis of protein folding suggests that loops of about 25 residues, closed through interactions between the loop ends (locks), play an important role in protein structure. Coarse-grain elastic network simulations, and examination of loop lengths in a diverse set of proteins, each supports a bias towards loops of close to 25 residues in length between residues of high stability. Previous studies have established a correlation between total contact distance (TCD), a metric of sequence distances between contacting residues (cf. contact order), and the log-folding rate of a protein. In a set of 43 proteins, we identify an improved correlation (r2 = 0.76), when the metric is restricted to residues contacting the locks, compared to the equivalent result when all residues are considered (r2 = 0.65). This provides qualified support for the hypothesis, albeit with an increased emphasis upon the importance of a much larger set of residues surrounding the locks. Evidence of a similar-sized protein core/extended nucleus (with significant overlap) was obtained from TCD calculations in which residues were successively eliminated according to their hydrophobicity and connectivity, and from molecular dynamics simulations. Our results suggest that while folding is determined by a subset of residues that can be predicted by application of the closed-loop hypothesis, the original hypothesis is too simplistic; efficient protein folding is dependent on a considerably larger subset of residues than those involved in lock formation.

Identifying subset errors in multiple sequence alignments

Multiple sequence alignment (MSA) accuracy is important, but there is no widely accepted method of judging the accuracy that different alignment algorithms give. We present a simple approach to detecting two types of error, namely block shifts and the misplacement of residues within a gap. Given a MSA, subsets of very similar sequences are generated through the use of a redundancy filter, typically using a 70–90% sequence identity cut-off. Subsets thus produced are typically small and degenerate, and errors can be easily detected even by manual examination. The errors, albeit minor, are inevitably associated with gaps in the alignment, and so the procedure is particularly relevant to homology modelling of protein loop regions. The usefulness of the approach is illustrated in the context of the universal but little known [K/R]KLH motif that occurs in intracellular loop 1 of G protein coupled receptors (GPCR); other issues relevant to GPCR modelling are also discussed.

The statistical significance of selected sense–antisense peptide interactions

Sense and antisense peptides, encoded by sense and corresponding antisense DNA strands, are capable of specific interactions that could be a driving force to mediate protein–protein or protein–peptide binding associations. The complementary residue hypothesis suggests that these interactions are founded upon the sum of pairwise interactions between amino acids encoded by corresponding sense and antisense codons. Despite many successful experimental results obtained with the hypothesis, however, the physicochemical basis for these interactions is poorly understood. We examined the potential of the hypothesis for general identification of protein–protein interaction sites, and the possible role of the hypothesis in determining folding in a broad set of protein structures. In addition, we performed a structural study to investigate the binding of a complementary peptide to IL‐1F2. Our results suggest that complementary residue pairs are no more frequent or conserved than average in protein–protein interfaces, and are statistically under‐represented amongst contacting residue pairs in folded protein structures. Although our structural results matched experimental observations of binding between the peptide and IL‐1F2, complementary residue interactions do not appear to be dominant in the bound structure. Overall, our data do not allow us to conclude that the complementary residue hypothesis accounts for specific sense–antisense peptide interactions.

The statistical significance of selected sense–antisense peptide interactions [J. Comp. Chem. 33, 1440–1447]

[a] C. J. R. Illingworth, S. V. Chintapalli, C. A. Reynolds Department of Biological Sciences, University of Essex, Wivenhoe Park, Colchester, CO4 3SQ, United Kingdom E-mail: reync@essex.ac.uk [b] S. A. Serapian, A. D. Miller Imperial College Genetic Therapies Centre, Department of Chemistry, Imperial College London, London SW7 2AZ, United Kingdom [c] A. D. Miller, V. Veverka, M. D. Carr Institute of Pharmaceutical Science, Kings College London, London SE1 9NH, United Kingdom [d] V. Veverka, M. D. Carr Department of Biochemistry, University of Leicester, Leicester LE1 9HN, United Kingdom Present address: Institute of Organic Chemistry and Biochemistry, Flemingovo nam. 2, Prague, Czech Republic.

Quantitative measurement of protease ligand conformation

The tendency for protease ligands to bind in an extended conformation has been suggested as an important factor for the identification of compounds of medicinal importance. Here we present a novel graph-theoretical method giving a quantitative measure of ligand conformation, and through application of this method to a representative set of protease ligands in bound and unbound conformations, derive the result that protease ligands are more extended in conformation when in their bound state.