Parametrically Managed Activation Functions for Improved Global Potential Energy Surfaces for Six Coupled 5A' States and Fourteen Coupled 3A' States of O + O2.

We report new potential energy surfaces for six coupled 5A' states and 14 coupled 3A' states of O3. The new surfaces are created by parametrically managed diabatization by deep neural network (PM-DDNN). The PM-DDNN method uses calculated adiabatic potential energy surfaces to discover and fit an underlying adiabatic-equivalent set of diabatic surfaces and their couplings and obtains the fit to the adiabatic surfaces by diagonalization of the diabatic potential energy matrix (DPEM). The procedure yields the adiabatic surfaces and their gradients, as well as the DPEM and its gradient. If desired one can also compute the nonadiabatic coupling due to the transformation. The present work improves on previous work by using a new coordinate to guide the decay of the neural network contribution to the many-body fit to the whole DPEM. The main objective was to obtain smoother potentials than the previous ones with better suitability for dynamics calculations, and this was achieved. Furthermore, we obtained suitably small deviations from the input reference data. For the six coupled 5A' surfaces, the 60,366 data below 10 eV are fit with a mean unsigned error (MUE) of 49 meV, and for the 14 coupled 3A' surfaces, the 76,733 data below 10 eV are fit with an MUE of 28 meV. The data below 5 eV fit even more accurately with MUEs of 37 meV (5A') and 20 meV (3A').

Evolutionary analysis reveals the origin of sodium coupling in glutamate transporters.

Secondary active membrane transporters harness the energy of ion gradients to concentrate their substrates. Homologous transporters evolved to couple transport to different ions in response to changing environments and needs. The bases of such diversification, and thus principles of ion coupling, are unexplored. Employing phylogenetics and ancestral protein reconstruction, we investigated sodium-coupled transport in prokaryotic glutamate transporters, a mechanism ubiquitous across life domains and critical to neurotransmitter recycling in humans. We found that the evolutionary transition from sodium-dependent to independent substrate binding to the transporter preceded changes in the coupling mechanism. Structural and functional experiments suggest that the transition entailed allosteric mutations, making sodium binding dispensable without affecting ion-binding sites. Allosteric tuning of transporters' energy landscapes might be a widespread route of their functional diversification.

Semiclassical Multistate Dynamics for Six Coupled 5A' States of O + O2.

Dynamics simulations of high-energy O2-O collisions play an important role in simulating thermal energy content and heat flux in flows around hypersonic vehicles. To carry out such dynamics simulations efficiently requires accurate global potential energy surfaces and (in most algorithms) state couplings for many energetically accessible electronic states. The ability to treat collisions involving many coupled electronic states has been a challenge for decades. Very recently, a new diabatization method, the parametrically managed diabatization by deep neural network (PM-DDNN), has been developed. The PM-DDNN method uses a deep neural network architecture with an activation function parametrically dependent on input data to discover and fit the diabatic potential energy matrix (DPEM) as a function of geometry, and the adiabatic potential energy surfaces are obtained by diagonalization of a small matrix with analytic matrix elements. Here, we applied the PM-DDNN method to the six lowest-energy potential energy surfaces in the 5A' manifold of O3 to perform simultaneous diabatization and fitting; the data are obtained by extended multistate complete-active-space second-order perturbation theory. We then used the adiabatic surfaces for dynamics calculations with three methods: coherent switching with decay of mixing (CSDM), curvature-driven CSDM (κCSDM), and electronically curvature-driven CSDM (eκCSDM). The κCSDM calculations require only adiabatic potential energies and gradients. The three dynamical methods are in good agreement. We then calculated electronically nonadiabatic, electronically inelastic, and dissociative cross sections for seven initial collision energies, five initial vibrational levels, and four initial rotational levels. Trends in the electronically inelastic cross sections as functions of the initial collision energy and vibrational level were rationalized in terms of the coordinate ranges where the gaps between the second and third potential energy surfaces are small.

Parametrically Managed Activation Function for Fitting a Neural Network Potential with Physical Behavior Enforced by a Low-Dimensional Potential.

Machine-learned representations of potential energy surfaces generated in the output layer of a feedforward neural network are becoming increasingly popular. One difficulty with neural network output is that it is often unreliable in regions where training data is missing or sparse. Human-designed potentials often build in proper extrapolation behavior by choice of functional form. Because machine learning is very efficient, it is desirable to learn how to add human intelligence to machine-learned potentials in a convenient way. One example is the well-understood feature of interaction potentials that they vanish when subsystems are too far separated to interact. In this article, we present a way to add a new kind of activation function to a neural network to enforce low-dimensional constraints. In particular, the activation function depends parametrically on all of the input variables. We illustrate the use of this step by showing how it can force an interaction potential to go to zero at large subsystem separations without either inputting a specific functional form for the potential or adding data to the training set in the asymptotic region of geometries where the subsystems are separated. In the process of illustrating this, we present an improved set of potential energy surfaces for the 14 lowest 3A' states of O3. The method is more general than this example, and it may be used to add other low-dimensional knowledge or lower-level knowledge to machine-learned potentials. In addition to the O3 example, we present a greater-generality method called parametrically managed diabatization by deep neural network (PM-DDNN) that is an improvement on our previously presented permutationally restrained diabatization by deep neural network (PR-DDNN).

Deciphering the Mechanism of Binding Selectivity of Chlorofluoroacetamide-Based Covalent Inhibitors toward L858R/T790M Resistance Mutation.

Covalent modification of the oncogenic mutant epidermal growth factor receptor (EGFR) by small molecules is an efficient strategy for achieving an enhanced and sustained pharmacological effect in the treatment of non-small-cell lung cancer. NSP-037 (18), an irreversible inhibitor of the L858R/T790M double-mutant EGFR (EGFRDM) using α-chlorofluoroacetamide (CFA) as a novel warhead, has seven times the inhibition selectivity for EGFRDM over the wild type (EGFRWT), as compared to clinically approved osimertinib (7). Here, we employ multiple computational approaches to elucidate the mechanism underlining this improved selectivity, as well as the effect of CFA on the selectivity enhancement of inhibitor 18 over 7. We find that EGFRDM undergoes significantly larger conformational changes than EGFRWT upon binding to 18. The conformational stability of the diamine side chain and the CFA motif of 18 in the orthosteric site of EGFRDM is identified as key for the disparate binding mechanism and inhibitory prowess of 18 with respect to EGFRWT and EGFRDM and 18's higher selectivity than 7. The binding free energy of the 18-bound complexes is -6.38 kcal/mol greater than that of the 7-bound complexes, explaining the difference in selectivity of these inhibitors. Further, free energy decomposition analysis indicates that the electrostatic contribution of key residues plays an important role in the 18-bound complexes. QM/MM calculations show that the most favored mechanism for the Cys797 alkylation reaction is the direct displacement mechanism through a CFA-based inhibitor, producing a reaction with the lowest energy barrier and most stable product.

Molecular Mechanism Exploration of Potent Fluorinated PI3K Inhibitors with a Triazine Scaffold: Unveiling the Unusual Synergistic Effect of Pyridine-to-Pyrimidine Ring Interconversion and CF3 Defluorination.

The phosphatidylinostitol-3-kinase (PI3K)/AKT/mammalian target of rapamycin signaling pathway is a vital regulator of cell proliferation, growth, and survival, which is frequently overactivated in many human cancers. To this effect, PI3K, which is an important mediator of this pathway, has been pinpointed as a crucial target in cancer therapy and hence the importance of PI3K inhibitors. It was recently reported that defluorination and pyridine-to-pyrimidine ring interconversion increase the potency of specific small-molecule inhibitors of PI3K. Compound 4, an inhibitor with the difluorinated pyrimidine motif, was found to be eight times more potent against PI3K than compound 1, an inhibitor with the trifluorinated pyridine motif. This observation presents the need to rationally resolve the differential inhibitory mechanisms exhibited by both compounds. In this present work, we employed multiple computational approaches to investigate and distinguish the binding modes of 1 and 4 in addition to the effects they mediate on the secondary structure of PI3K. Likewise, we evaluated two other derivatives, compounds 2 with the difluorinated pyridine motif and 3 with the trifluorinated pyrimidine motif, to investigate the cooperativity effect between the defluorination of CF3 and pyridine-to-pyrimidine ring interconversion. Findings revealed that PI3K, upon interaction with 4, exhibited a series of structural changes that favored the binding of the inhibitor at the active-site region. Furthermore, a positive (synergistic) cooperativity effect was observed between CF3 defluorination and pyridine-to-pyrimidine ring interconversion. Moreover, there was a good correlation between the binding free energy estimated and the biological activity reported experimentally. Energy decomposition analysis revealed that the major contributing force to binding affinity variations between 1 and 4 is the electrostatic energy. Per-residue energy-based hierarchical clustering analysis further identified four hot-spot residues ASP841, TYR867, ASP964, and LYS833 and four warm-spot residues ASP836, SER806, ASP837, and LYS808, which essentially mediate the optimal and higher-affinity binding of compound 4 to PI3K relative to 1. This study therefore provides rational insights into the mechanisms by which 4 exhibited superior PI3K-inhibitory activities over 1, which is vital for future structure-based drug discovery efforts in PI3K targeting.

Mechanistic Study of Potent Fluorinated EGFR Kinase Inhibitors with a Quinazoline Scaffold against L858R/T790M/C797S Resistance Mutation: Unveiling the Fluorine Substituent Cooperativity Effect on the Inhibitory Activity.

Fluorination has considerable potential with regard to the design of kinase inhibitors for anticarcinoma therapy. It was recently reported that fluorination increases the potency of inhibitors of the epidermal growth factor receptor (EGFR), mutations of which have been linked specifically to nonsmall-cell lung cancer. For the L858R/T790M/C797S triplet mutant (EGFRTM), a difluorinated inhibitor, 25g, was found to have 4.23 times greater potency against the EGFRTM than an unfluorinated inhibitor, 25a. This discovery necessitates a rational explanation for the underlying inhibitory mechanisms. Here, we apply multiple computational approaches to explore, validate, and differentiate the binding modes of 25a and 25g in the EGFRTM and investigate the cooperativity effect of fluorine substituents on the inhibitory activity. Our results showed that the EGFRTM in the presence of 25g undergoes a series of conformational changes that favor inhibitor binding to both the active and allosteric sites. Further, the cooperativity effect of fluorine substituents is positive: the complex stability is increased by each additional fluorine substituent. Estimated binding free energies show good correlation with the experimental biological activity. Subsequently, the decomposition energy analysis revealed that the van der Waals interaction is the principal force contributing to variations in the binding affinities of 25a and 25g to the EGFRTM. Per-residue energy-based hierarchical clustering analysis suggests that three hot-spot residues, L718, K745, and D855, are the key in achieving optimal binding modes for 25g with higher affinity in the EGFRTM compared to 25a. This study provides a rationale for the superior EGFRTM-inhibitory potency exhibited by 25g over 25a, which is expected to be useful for the future rational structure-based design of novel EGFRTM inhibitors with improved potency and selectivity.

A Mechanistic Study of a Potent and Selective Epidermal Growth Factor Receptor Inhibitor against the L858R/T790M Resistance Mutation.

Covalent targeting is a promising strategy for increasing the potency and selectivity of potential drug candidates. This therapeutic approach was recently reported for the epidermal growth factor receptor (EGFR), wherein a covalent binder, 20g [N-(3-{7-[2-methoxy-4-(4-methylpiperazin-1-yl)phenylamino]-3,4-dihydro-3-isopropyl-2,4-dioxopyrimido[4,5-d]pyrimidin-1(2H)-yl}phenyl)acrylamide], demonstrated significant selectivity and inhibitory activity toward the EGFR L858R/T790M double mutant (EGFRDM) relative to the EGFR wild-type form (EGFRWT). The enhanced therapeutic potency of 20g against EGFRDM is 263 times greater than that against EGFRWT, which necessitates a rational explanation for the underlying selective and inhibitory mechanisms. In this work, we investigate the differential binding modes of 20g with EGFRWT and EGFRDM using molecular dynamics simulations coupled with free energy calculations and further identify key residues involved in the selective targeting, binding, and inhibitory mechanisms mediated by 20g. We find that systematic orientational and conformational changes in the α-loop, p-loop, active loop, and αC-helix are responsible for the disparate binding mechanisms and inhibitory prowess of 20g with respect to EGFRWT and EGFRDM. The calculated binding free energies show good correlation with the experimental biological activity. The total binding free energy difference between EGFRWT-20g and EGFRDM-20g is -11.47 kcal/mol, implying that 20g binds more strongly to EGFRDM. This enhanced binding affinity of 20g for EGFRDM is a result of a large increase in the van der Waals and electrostatic interactions with three critical residues (Met790, Gln791, and Met793) that are chiefly responsible for the high-affinity interactions mediated by 20g with EGFRDM relative to EGFRWT.

Discovery of novel natural flavonoids as potent antiviral candidates against hepatitis C virus NS5B polymerase.

The non-structural 5B (NS5B) polymerase of hepatitis C virus (HCV) is an attractive target for antiviral intervention. Quercetagetin (Que) is a natural flavonoid, which has been exhibited to have anti-HCV property through inhibition of RNA binding to NS5B. The last few decades have witnessed a growing interest in the extraction of natural flavonoids with a plethora of different biological activities. Considering the high therapeutic potential of Que, the aim of this study is to explore wide structure entities with potent activity using Que as a prototype. A virtual screen protocol involving docking and molecular dynamics has been performed to examine the potency of forty-three natural flavonoids which recently extracted from plants for inhibition of NS5B. During two screening stages, two compounds 24 and 41 were identified to have more favorable binding affinity to NS5B as compared to Que. The comparative analysis showed that there is a significant difference in the binding free energy of Que and 41 (ΔΔGbind = -11.17 kcal/mol). It was revealed that van der Waals (vdW) interaction drives the binding process of both 24 and 41 and plays an important role in increasing their activities relative to Que. PHE162 serves as a crucial residue in both the NS5B-24 and NS5B-41 systems, contributing the most vdW energy by π-π interaction, suggesting that aromatic interactions are critical for the binding of 24 and 41 to NS5B. Moreover, hydrogen bond analysis indicates that the hydrogen bonds formed by LYS98, THR137, ASP164 and ARG168, can play important roles in the increased binding affinity of 41 to NS5B relative to Que. The findings of this study will provide useful structure-activity relationship (SAR) guidelines for the design of novel inhibitors with improved/enhanced therapeutic activities in the treatment of hepatitis C.

Revealing the distinct mechanistic binding and activity of 5-(1-(3,5-dichloropyridin-4-yl)ethoxy)-3-(5-(4-methylpiperazin-1-yl)-1H-benzo[d]imidazol-2-yl)-1H-indazole enantiomers against FGFR1.

The concept of chirality has become prominent over the years, particularly with regards to the design of therapeutic molecules. This phenomenon was recently reported for pro-carcinogenic fibroblast growth factor receptor 1 (FGFR1), wherein two inhibitors exhibited disparate inhibitory potencies due to the effects of chirality. Therefore, the ability of the R-enantiomer (R-21c) to possess a potency 10.44 times that of the S-enantiomer (S-21c) leaves us with a curiosity to investigate the underlying mechanisms using computational methods. Hence, presented in this study are insights that clearly explain the systematic effects of chirality on the differential activities of (R)-21c and (S)-21c towards FGFR1. The findings showed that the "R-configured" (R)-21c induced a notable conformational change in the active site P-loop, which enhanced its motion, as complemented by rotation of two dihedral angles: φ1(CNCC) and φ2(CC*OC), providing a favorable orientation. Likewise, optimal positioning of (R)-21c at the binding cavity allowed adequate interspaces that facilitated the formation of strong interactions with target residues. Moreover, the estimated ΔG binding correlated with bioactivity data (IC50) and, when decomposed, we observed that van der Waals (vdW) interactions were the major highlight of the binding process of both 21c enantiomers and also accounted for their differential activities. Active site interactions of (R)-21c with residues Phe489 and Arg629 stabilized its two benzimidazole motifs, while Arg570 and Pro663 formed two strong NH-N hydrogen bonds and one π-alkyl interaction, which altogether accounted for its inhibitory prowess towards FGFR1. In contrast, these interactions were not observed in (S)-21c due to its non-flexible S-configuration, which disallowed its extension into the active site region and prevented interaction with crucial residues. These results are expected to facilitate the discovery and rational design of novel and specific FGFR1 inhibitors.

Probing the Dynamic Mechanism of Uncommon Allosteric Inhibitors Optimized to Enhance Drug Selectivity of SHP2 with Therapeutic Potential for Cancer Treatment.

There is currently considerable interest in SHP2 as a potential target for treatment of cancer. Mutation in SHP2, particularly the E76A mutation, has been found to seriously confer the phosphatase high activity. Recently, two compounds, 1 and 23, have been reported as potent allosteric inhibitors of both SHP2 wild type (SHP2WT) and the E76A mutant (SHP2E76A), with higher activity than other inhibitors. However, the structural and dynamic implications of their inhibitory mechanisms are yet unexplored which deserve further attention. Herein, the MD simulation applies to gain insight into the atomistic nature of each binding mode of inhibitors 1 and 23 in both SHP2WT and SHP2E76A. The comparative analysis reveals inhibitor 1 can freeze SHP2WT and SHP2E76A in their auto-inhibited conformation better than 23, in agreement with experimental data. GLU250 in both SHP2WT and SHP2E76A and ARG111 and ARG229 in SHP2E76A play a crucial role in the higher activity of 1 compared to 23. The mutation E76A increases the binding affinity of 1 and 23 compared to the wild type, implying that the two inhibitors have been well adopted by the E76A mutant. The findings here can substantially shed light on new strategies for developing novel classes of SHP2 inhibitors with increased potency.

Covalent vs. Non-Covalent Inhibition: Tackling Drug Resistance in EGFR - A Thorough Dynamic Perspective.

A persistent challenge in the treatment of non-small cell lung cancer (NSCLC) with EGFR is the emergence of drug-resistant caused by somatic mutations. The EGFR L858R/T790 M double mutant (EGFRDM ) was found to be the most alarming variant. Despite the development of a wide range of inhibitors, none of them could inhibit EGFRDM effectively. Recently, 11h and 45a, have been found to be potent inhibitors against EGFRDM through two distinctive mechanisms, non-covalent and covalent binding, respectively. However, the structural and dynamic implications of the two modes of inhibitions remain unexplored. Herein, two molecular dynamics simulation protocols, coupled with free-energy calculations, were applied to gain insight into the atomistic nature of each binding mode. The comparative analysis confirmed that there is a significant difference in the binding free energy between 11h and 45a (ΔΔGbind =-21.17 kcal/mol). The main binding force that governs the binding of both inhibitors is vdW, with a higher contribution for 45a. Two residues ARG841 and THR854 were found to have curtailed role in the binding of 45a to EGFRDM by stabilizing its flexible alcohol chain. The 45a binding to EGFRDM induces structural rearrangement in the active site to allow easier accessibility of 45a to target residue CYS797. The findings of this work can substantially shed light on new strategies for developing novel classes of covalent and non-covalent inhibitors with increased specificity and potency.

Noteworthy effect of slight variation in aliphatic chain length of trisubstituted imidazole inhibitors against epidermal growth factor receptor L858R/T790M/C797S mutant in cancer therapy.

11h is a very potent inhibitor against epidermal growth factor receptor triple mutation L858R/T790M/C797S (EGFRTM ) with 13-fold stronger potency than the FDA-approved osimertinib. Recently, two new EGFRTM inhibitors, 11d and 11e, were reported which revealed 2.8- and 2.3-fold stronger potency than 11h, respectively. 11h, 11d, and 11e have the same structures but differ only in their aliphatic chain length. However, the exact effects of differential aliphatic chain length on the inhibitory potencies of these compounds require further elaboration at the atomistic level, hence the objective of this report. Various computational tools were employed for this purpose. From our findings, it was revealed that van der Waals (vdW) interactions modulate the binding mechanisms of these inhibitors and play the most important role in the differential inhibitory activities of 11d, 11h, and 11e. The strong hydrogen bond formation between the aliphatic chain of 11d and key residue ARG841 was recognized as the reason for its higher activity and inhibitory potency relative to 11h and 11e. Moreover, the extension of the N-terminal loop into the active site for vdW interaction with the phenyl group of 11e and carbon-hydrogen bond formed between the aliphatic chain of 11e and LEU718 engendered a higher activity of 11e than 11h.

A theoretical study of π-stacking interactions in C-substituted tetrazoles.

The π-stacking effects of benzene ring (Ben) with 1H- and 2H-tetrazole derivatives (1H-TZ-X and 2H-TZ-X) substituted at C5 (where X is Cl, COH, NO, NO2, CN, NH2, OH, OCH3, SH and H) has been investigated by the quantum mechanical calculations at the M06-2X/6-311++G** level. The results indicate the 1H-TZ-X||Ben complexes (|| donates π-stacking interaction) are more stable than 2H-TZ-X||Ben while in unstacked forms, 1H-TZ-X is less stable than 2H-TZ-X. All substituents enhance the π-stacking interaction relative to the unsubstituted ones and enhancement is higher for the electron-withdrawing substituents (EWSs). Also, investigation of the local and direct effect of substituents in stacking interaction showed that all substituents regardless of whether are electron donating or electron withdrawing have an additive effect in π-stacking interaction. Excellent correlations were found between the binding energies of the complexes and combination of substituent constant terms. The results showed that the electrostatic interaction alone is not responsible for stacking stabilization but charge penetration is important. Furthermore, analysis of aromaticity, AIM, ESP and NPA were investigated to obtain aromaticity index, non-bonding interactions, chemical reactivity and polarity (dipole moment), respectively.

π-Stacking effects on the hydrogen bonding capacity of methyl 2-naphthoate.

π-stacking effects of fused two-ring system of methyl 2-naphthoate (MNP) with benzene derivatives on the CO group, as a hydrogen bond acceptor, has been investigated by the quantum mechanical calculations at the M06-2X/6-311++G(d,p) level of theory. All substituents enhance the stacking interactions relative to the unsubstituted case, where enhancement is higher for electron-withdrawing substituents (EWSs). The hydrogen bonding ability of lone pairs of O* atom of stacked MNP decreases in the presence of strong electron-withdrawing substituents (NO2, NO and CN). The hydrogen bond ability of CO group of MNP is related to the sum of local minima of electrostatic potentials (∑ESPs) observed between stacked rings. The charge transfer (CT) is lower in the presence of EWSs. The study also shows that the interaction energies (ΔE) are linearly dependent on the combination of the sum of electron densities calculated at the bond critical points (BCPs) between the rings (∑ρBCP) and the sum of electron charge densities calculated at the ring critical points (∑ρRCP). There are good relationships between the Hammett constant σmeta and the global minimum of electrostatic potential around the O* atom (Vmin), the sum of local minima of the electrostatic potentials obtained between stacked rings, and the results of natural population analysis (NPA). An excellent correlation was found between the ΔE values and a combination of the electrostatic (σmeta), resonance/induction (σpara) and dispersion/polarizibility (molar refractivity, MR) substituent constant terms.

Caffeine as base analogue of adenine or guanine: a theoretical study.

The results of quantum mechanical calculations, including binding energies and results of the population analysis show that the GC and AT base pair complexes are more stable than the CAF-X ones (where CAF is caffeine and X=adenine (A), thymine (T), cytosine (C) and guanine (G)). Structural similarity between the CAF molecule and purine bases (G and A) provides the possibility of incorporation of the CAF molecule into the DNA macromolecule. By comparing the CAF-A and CAF-T complexes with the AT base pair, and the CAF-G and CAF-C complexes with the GC base pair, it was found that the formation of the CAF-T complex is more probable than the other complexes. Thus, the CAF molecule acts as an analogue base of A and can be incorporated into the DNA macromolecule and paired with T and C in normal and rare state, respectively. Indeed, the results show that formation of the CAF-C complex is less probable than the CAF-T one and an AT to GC conversion is rarely occurred in the next DNA replication, so the CAF molecule may be considered as a weak mutagenic compound. To examine solvent effect, the binding energies have been calculated in solvent for the most important structures of the CAF-G, CAF-T, CAF-A and CAF-C complexes. The results in solvent are in agreement with those in the gas phase.