Charges for Large Scale Binding Free Energy Calculations with the Linear Interaction Energy Method.

The linear interaction energy method (LIE), which combines force field based molecular dynamics (MD) simulations and linear response theory, has previously been shown to give fast and reliable estimates of ligand binding free energies, suggesting that this type of technique could be used also in a high-throughput fashion. However, a limiting step in such applications is the assignment of atomic charges for compounds that have not been parametrized within the given force field, in this case OPLS-AA. In order to reach an automatable solution to this problem, we have examined the performance of nine different ab initio and semiempirical charge methods, together with estimates of solvent induced polarization. A test set of ten HIV-1 reverse transcriptase inhibitors was selected, and LIE estimates of their relative binding free energies were calculated using the resulting 23 different charge variants. Over 800 ns of MD simulation show that the LIE method provides excellent estimates with several different charge methods and that the semiempirically derived CM1A charges, in particular, emerge as a fast and reliable alternative for fully automated LIE based virtual screens with the OPLS-AA force field. Our conclusions regarding different charge models are also expected to be valid for other types of force field based binding free energy calculations, such as free energy perturbation and thermodynamic integration simulations.

Predicting binding modes from free energy calculations.

To produce reliable predictions of bioactive conformations is a major challenge in the field of structure-based inhibitor design and is a requirement for accurate binding free energy predictions with structure-based methods. A series of HIV-1 reverse transcriptase inhibitors was cross-docked using a non-native crystal structure that resulted in two distinct clusters of possible conformations. One of these clusters was compatible with an existing crystal structure, whereas the other displayed a flipped heterocyclic group. Binding free energies, using the non-native crystal structure, calculated from several scoring functions, were similar for the two clusters, and no conclusion about the binding mode could be drawn from these results. The two clusters could be separated through rescoring with the linear interaction method (LIE) in combination with molecular dynamics simulations, which leads to a binding mode prediction in line with experimental crystallographic data. Further, the LIE model produces the best correlation between experimental and calculated binding free energies among the tested scoring methods.

Computational analysis of plasmepsin IV bound to an allophenylnorstatine inhibitor.

The plasmepsin proteases from the malaria parasite Plasmodium falciparum are attracting attention as putative drug targets. A recently published crystal structure of Plasmodium malariae plasmepsin IV bound to an allophenylnorstatine inhibitor [Clemente, J.C. et al. (2006) Acta Crystallogr. D 62, 246-252] provides the first structural insights regarding interactions of this family of inhibitors with plasmepsins. The compounds in this class are potent inhibitors of HIV-1 protease, but also show nM binding affinities towards plasmepsin IV. Here, we utilize automated docking, molecular dynamics and binding free energy calculations with the linear interaction energy LIE method to investigate the binding of allophenylnorstatine inhibitors to plasmepsin IV from two different species. The calculations yield excellent agreement with experimental binding data and provide new information regarding protonation states of active site residues as well as conformational properties of the inhibitor complexes.

Computational study of the binding affinity and selectivity of the bacterial ammonium transporter AmtB.

We report results from microscopic molecular dynamics and free energy perturbation simulations of substrate binding and selectivity for the Escherichia coli high-affinity ammonium transporter AmtB. The simulation system consists of the protein embedded in a model membrane/water surrounding. The calculated absolute binding free energies for the external NH(4)(+) ions are between -5.8 and -7.3 kcal/mol and are in close agreement with experimental data. The apparent pK(a) of the bound NH(4)(+) increases by more than 4 units, indicating a preference for binding ammonium ion and not neutral ammonia. The external binding site is also selective for NH(4)(+) toward monovalent metal cations by 2.4-4.4 kcal/mol. The externally bound NH(4)(+) shows strong electrostatic interactions with the proximal buried Asp160, stabilized in the anionic form, whereas the interactions with the aromatic rings of Phe107 and Trp148, lining the binding cavity, are less pronounced. Simulated mutation of the highly conserved Asp160 to Asn reduces the pK(a) of the bound ammonium ion by approximately 7 units and causes loss of its binding. The calculations further predict that the substrate affinity of E. coli AmtB depends on the ionization state of external histidines. The computed free energies of hypothetical intermediate states related to transfer of NH(3), NH(4)(+), or H(2)O from the external binding site to the first position inside the internal channel pore favor permeation of the neutral species through the channel interior. However, the predicted change in the apparent pK(a) of NH(4)(+) upon translocation from the external site, Am1, to the first internal site, Am2, indicates that ammonium ion becomes deprotonated only when it enters the channel interior.

Inhibitor binding to the plasmepsin IV aspartic protease from Plasmodium falciparum.

Plasmepsin IV (Plm IV) is one of the aspartic proteases present in the food vacuole of the malaria parasite Plasmodium falciparum involved in host hemoglobin degradation by the parasite. Using a series of previously synthesized plasmepsin inhibitors [Ersmark, K., et al. (2005) J. Med. Chem. 48, 6090-106], we report here experimental data and theoretical analysis of their inhibitory activity toward Plm IV. All compounds share a 1,2-dihydroxyethylene unit as the transition state mimic. They possess symmetric P1 and P1' side chains and either a diacylhydrazine, a five-membered oxadiazole ring, or a retroamide at the P2 and P2' positions. Experimental binding affinities are compared to those predicted by the linear interaction energy (LIE) method and an empirical scoring function, using both a crystal structure and a homology model for the enzyme. Molecular dynamics (MD) simulations of the modeled complexes allow a rational interpretation of the structural determinants for inhibitor binding. A ligand bearing a P2 and P2' symmetric oxadiazole which is devoid of amide bonds is identified both experimentally and theoretically as the most potent inhibitor of Plm IV. For the P2 and P2' asymmetric compounds, the results are consistent with earlier predictions regarding the mode of binding of this class of inhibitors to Plm II. Theoretical estimation of selectivity for some compounds is also reported. Significant features of the Plm IV binding pocket are discussed in comparison to related enzymes, and the results obtained here should be helpful for further optimization of inhibitors.

Continuum solvation models in the linear interaction energy method.

The linear interaction energy (LIE) method in combination with two different continuum solvent models has been applied to calculate protein-ligand binding free energies for a set of inhibitors against the malarial aspartic protease plasmepsin II. Ligand-water interaction energies are calculated from both Poisson-Boltzmann (PB) and Generalized Born (GB) continuum models using snapshots from explicit solvent simulations of the ligand and protein-ligand complex. These are compared to explicit solvent calculations, and we find close agreement between the explicit water and PB solvation models. The GB model overestimates the change in solvation energy, and this is caused by consistent underestimation of the effective Born radii in the protein-ligand complex. The explicit solvent LIE calculations and LIE-PB, with our standard parametrization, reproduce absolute experimental binding free energies with an average unsigned error of 0.5 and 0.7 kcal/mol, respectively. The LIE-GB method, however, requires a constant offset to approach the same level of accuracy.

The molecular mechanism of peptide-mediated erythromycin resistance.

The macrolide antibiotic erythromycin binds at the entrance of the nascent peptide exit tunnel of the large ribosomal subunit and blocks synthesis of peptides longer than between six and eight amino acids. Expression of a short open reading frame in 23 S rRNA encoding five amino acids confers resistance to erythromycin by a mechanism that depends strongly on both the sequence and the length of the peptide. In this work we have used a cell-free system for protein synthesis with components of high purity to clarify the molecular basis of the mechanism. We have found that the nascent resistance peptide interacts with erythromycin and destabilizes its interaction with 23 S rRNA. It is, however, in the termination step when the pentapeptide is removed from the peptidyl-tRNA by a class 1 release factor that erythromycin is ejected from the ribosome with high probability. Synthesis of a hexa- or heptapeptide with the same five N-terminal amino acids neither leads to ejection of erythromycin nor to drug resistance. We propose a structural model for the resistance mechanism, which is supported by docking studies. The rate constants obtained from our biochemical experiments are also used to predict the degree of erythromycin resistance conferred by varying levels of resistance peptide expression in living Escherichia coli cells subjected to varying concentrations of erythromycin. These model predictions are compared with experimental observations from growing bacterial cultures, and excellent agreement is found between theoretical prediction and experimental observation.

Macrocyclic inhibitors of the malarial aspartic proteases plasmepsin I, II, and IV.

The first macrocyclic inhibitor of the Plasmodium falciparum aspartic proteases plasmepsin I, II, and IV with considerable selectivity over the human aspartic protease cathepsin D has been identified. A series of macrocyclic compounds were designed and synthesized. Cyclizations were accomplished using ring-closing metathesis with the second generation Grubbs catalyst. These compounds contain either a 13-membered or a 16-membered macrocycle and incorporate a 1,2-dihydroxyethylene as transition state mimicking unit. The binding mode of this new class of compounds was predicted with automated docking and molecular dynamics simulations, with an estimation of the binding affinities through the linear interaction energy (LIE) method.

Synthesis of malarial plasmepsin inhibitors and prediction of binding modes by molecular dynamics simulations.

A series of inhibitors of the malarial aspartic proteases Plm I and II have been synthesized with L-mannitol as precursor. These inhibitors are characterized by either a diacylhydrazine or a five-membered oxadiazole ring replacing backbone amide functionalities. Molecular dynamics simulations were applied in the design process. The computationally predicted Plm II Ki values were generally in excellent agreement with the biological results. The diacylhydrazine was found to be superior over the oxadiazole as an amide bond replacement in the Plm I and II inhibitors studied. An extensive flexibility of the S2' pocket was captured by the simulations predicting the binding mode of the unsymmetrical inhibitors. Plm I and II inhibitors with single digit nanomolar Ki values devoid of inhibitory activity toward human Cat D were identified. One compound, lacking amide bonds, was found to be Plm IV selective and very potent, with a Ki value of 35 nM.