Quantum Mechanical Prediction of Dissociation Constants for Thiazol-2-imine Derivatives.

As weak acids or bases, in solution, drug molecules are in either their ionized or nonionized states. A high degree of ionization is essential for good water solubility of a drug molecule and is required for drug-receptor interactions, whereas the nonionized form improves a drug's lipophilicity, allowing the ligand to cross the cell membrane. The penetration of a drug ligand through cell membranes is mainly governed by the pKa of the drug molecule and the membrane environment. In this study, with the aim of predicting the acetonitrile pKa's (pKa(MeCN)) of eight drug-like thiazol-2-imine derivatives, we propose a very accurate and computationally affordable protocol by using several quantum mechanical approaches. Benchmark studies were conducted on a set of training molecules, which were selected from the literature with known pKa(water) and pKa(MeCN). Highly well-correlated pKa values were obtained when the calculations were performed with the isodesmic method at the M062X/6-31G** level of theory in conjunction with SMD solvation model for nitrogen-containing heterocycles. Finally, experimentally unknown pKa(MeCN) values of eight thiazol-2-imine structures, which were previously synthesized by some of us, are proposed.

Unexplored Vinylic-Substituted 5-Benzylidenethiazolidine-2,4-diones: Synthesis and DFT/NMR Stereochemical Assignment.

By exploring an efficient and versatile method for the 6-functionalization of its scaffold, we investigated the opening of a new chemical space around benzylidenethiazolidine-2,4-dione (BTZD). The 6-chloro- and 6-formyl BTZD obtained in two steps starting from 5-lithioTZD were selected as key intermediates and involved in a Pd-catalyzed cross-coupling or Wittig olefination. A variety of aryl, heteroaryl, or alkenyl substituents was successfully introduced on the vinylic position of BTZD, and particular attention was paid to elucidate the stereochemistry of the benzylidene derivatives by using a combined DFT/NMR study.

Molecular Mechanism of Protein Arginine Deiminase 2: A Study Involving Multiple Microsecond Long Molecular Dynamics Simulations.

Peptidylarginine deiminase 2 (PAD2) is a Ca2+-dependent enzyme that catalyzes the conversion of protein arginine residues to citrulline. This kind of structural modification in histone molecules may affect gene regulation, leading to effects that may trigger several diseases, including breast cancer, which makes PAD2 an attractive target for anticancer drug development. To design new effective inhibitors to control activation of PAD2, improving our understanding of the molecular mechanisms of PAD2 using up-to-date computational techniques is essential. We have designed five different PAD2-substrate complex systems based on varying protonation states of the active site residues. To search the conformational space broadly, multiple independent molecular dynamics simulations of the complexes have been performed. In total, 50 replica simulations have been performed, each of 1 µs, yielding a total simulation time of 50 µs. Our findings identify that the protonation states of Cys647, Asp473, and His471 are critical for the binding and localization of the N-α-benzoyl-l-arginine ethyl ester substrate within the active site. A novel mechanism for enzyme activation is proposed according to near attack conformers. This represents an important step in understanding the mechanism of citrullination and developing PAD2-inhibiting drugs for the treatment of breast cancer.

Impact of Deamidation on the Structure and Function of Antiapoptotic Bcl-xL.

Bcl-xL is an antiapoptotic mitochondrial trans-membrane protein, which is known to play a crucial role in the survival of tumor cells. The deamidation of Bcl-xL is a pivotal switch that regulates its biological function. The potential impact of deamidation on the structure and dynamics of Bcl-xL is directly linked to the intrinsically disordered region (IDR), which is the main site for post-translational modifications (PTMs). In this study, we explored deamidation-induced conformational changes in Bcl-xL to gain insight into its loss of function by performing microsecond-long molecular dynamics (MD) simulations. MD simulation outcomes showed that the IDR motion and interaction patterns have changed notably upon deamidation. Principal component analysis (PCA) demonstrates significant differences between wild-type and deamidated Bcl-xL and suggests that deamidation affects the structure and dynamics of Bcl-xL. The combination of clustering analysis, H-bond analysis, and PCA revealed changes in conformation, interaction, and dynamics upon deamidation. Differences in contact patterns and essential dynamics that lead to a narrowing in the binding groove (BG) are clear indications of deamidation-induced allosteric effects. In line with previous studies, we show that the IDR plays a very important role in the loss of apoptotic functions of Bcl-xL while providing a unique perspective on the underlying mechanism of Bcl-xL deamidation-induced cell death.

Using Atomic Charges to Describe the pKa of Carboxylic Acids.

In this study, we present an accurate protocol for the fast prediction of pKa's of carboxylic acids based on the linear relationship between computed atomic charges of the anionic form of the carboxylate fragment and their experimental pKa values. Five charge descriptors, three charge models, three solvent models, gas-phase calculations, several DFT methods (a combination of eight DFT functionals and fifteen basis sets), and four different semiempirical approaches were tested. Among those, the best combination to reproduce experimental pKa's is to compute the natural population analysis atomic charge using the solvation model based on density model at the M06L/6-311G(d,p) level of theory and selecting the maximum atomic charge on the carboxylic oxygen atoms (R2 = 0.955). The applicability of the suggested protocol and its stability along geometrical changes are verified by molecular dynamics simulations performed for a set of aspartate, glutamate, and alanine peptides. By reporting the calculated atomic charge of the carboxylate form into the linear relationship derived in this work, it should be possible to accurately estimate the amino acid's pKa's in a protein environment.

Elucidation of the atroposelectivity in the synthesis of axially chiral thiohydantoin derivatives.

Recently, Sarigul and Dogan have synthesized a number of enantiomerically enriched axially chiral atropoisomeric 2-thiohydantoins by the reaction of l-amino acid ester salts and o-aryl isothiocyanates in the presence of triethyl amine (TEA) in dichloromethane. The non-axially chiral derivative 5-methyl-3-phenyl-2-thiohydantoin gave a racemic product whereas the axially chiral 5-methyl-3-o-bromophenyl-2-thiohydantoin was less prone to racemize at C5 of the heterocyclic ring. In this study, we present a computational study (M06-2X/6-311+G(d,p) for C, H, O, N and S; M06-2X/6-311++G(3df,3pd) for Br) in order to propose plausible mechanisms for the racemization and cyclization steps for 2-thiohydantoin derivatives. The study includes rationalization based on steric as well as the electrostatic effects to elucidate the epimerization differences at C5.

Dendrigraft of Poly-l-lysine as a Promising Candidate To Reverse Heparin-based Anticoagulants in Clinical Settings.

By using a combination of experimental and computational experiments, we demonstrated that a second-generation dendrigraft of poly-l-lysine neutralizes the anticoagulant activity of unfractionated heparin, low-molecular-weight heparin, and fondaparinux more efficiently than protamine does in human plasma, making this synthetic polymer a promising surrogate of this problematic protein in clinical settings.

Semi-Empirical Born-Oppenheimer Molecular Dynamics (SEBOMD) within the Amber Biomolecular Package.

Semi-empirical quantum methods from the neglect of differential diatomic overlap (NDDO) family such as MNDO, AM1, or PM3 are fast albeit approximate quantum methods. By combining them with linear scaling methods like the divide & conquer (D&C) method, it is possible to quickly evaluate the energy of systems containing hundreds to thousands of atoms. We here present our implementation in the Amber biomolecular package of a SEBOMD module that provides a way to run semi-empirical Born-Oppenheimer molecular dynamics. At each step of a SEBOMD, a fully converged self-consistent field (SCF) calculation is performed to obtain the semiempirical quantum potential energy of a molecular system encaged or not in periodic boundary conditions. We describe the implementation and the features of our SEBOMD implementation. We show the requirements to conserve the total energy in NVE simulations, and how to accelerate SCF convergence through density matrix extrapolation. Specific ways of handling periodic boundary conditions using mechanical embedding or electrostatic embedding through a tailored quantum Ewald summation is developed. The parallel performance of SEBOMD simulations using the D&C scheme are presented for liquid water systems of various sizes, and a comparison between the traditional full diagonalization scheme and the D&C approach for the reproduction of the structure of liquid water illustrates the potentiality of SEBOMD to simulate molecular systems containing several hundreds of atoms for hundreds of picoseconds with a quantum mechanical potential in a reasonable amount of CPU time.

Synthesis and evaluation of new designed multiple ligands directed towards both peroxisome proliferator-activated receptor-γ and angiotensin II type 1 receptor.

Because of the complex biological networks, many pathologic disorders fail to be treated with a molecule directed towards a single target. Thus, combination therapies are often necessary, but they have many drawbacks. An alternative consists in building molecules intended to interact with multiple targets, called designed multiple ligands. We followed such a strategy in order to treat metabolic syndrome, by setting up molecules directed towards both type 1 angiotensin II (AT1) receptor and peroxisome proliferator-activated receptor-γ (PPAR-γ). For this purpose, many molecules were prepared by merging both pharmacophores following three different strategies. Their ability to activate PPAR-γ and to block AT1 receptors were evaluated in vitro. This strategy led to the preparation of many new PPAR-γ activating and AT1 blocking molecules. Among them, some exhibited both activities, highlighting the convenience of this approach.

Biotinylation enhances the anticancer effects of 15d­PGJ2 against breast cancer cells.

15-Deoxy-∆12,14-prostaglandin J2 (15d­PGJ2) is a natural agonist of peroxisome proliferator-activated receptor Î³ (PPARγ) that displays anticancer activity. Various studies have indicated that the effects of 15d­PGJ2 are due to both PPARγ-dependent and -independent mechanisms. In the present study, we examined the effects of a biotinylated form of 15d­PGJ2 (b­15d­PGJ2) on hormone-dependent MCF­7 and triple­negative MDA­MB­231 breast cancer cell lines. b­15d­PGJ2 inhibited cell proliferation more efficiently than 15d­PGJ2 or the synthetic PPARγ agonist, efatutazone. b­15d­PGJ2 was also more potent than its non-biotinylated counterpart in inducing apoptosis. We then analyzed the mechanisms underlying this improved efficiency. It was found not to be the result of biotin receptor-mediated increased incorporation, since free biotin in the culture medium did not decrease the anti-proliferative activity of b­15d­PGJ2 in competition assays. Of note, b­15d­PGJ2 displayed an improved PPARγ agonist activity, as measured by transactivation experiments. Molecular docking analyses revealed a similar insertion of b­15d­PGJ2 and 15d­PGJ2 into the ligand binding domain of PPARγ via a covalent bond with Cys285. Finally, PPARγ silencing markedly decreased the cleavage of the apoptotic markers, poly(ADP-ribose) polymerase 1 (PARP­1) and caspase­7, that usually occurs following b­15d­PGJ2 treatment. Taken together, our data indicate that biotinylation enhances the anti-proliferative and pro-apoptotic activity of 15d­PGJ2, and that this effect is partly mediated via a PPARγ-dependent pathway. These results may aid in the development of novel therapeutic strategies for breast cancer treatment.

Digitizing Poly-l-lysine Dendrigrafts: From Experimental Data to Molecular Dynamics Simulations.

Despite the growing use of poly-l-lysine dendrigrafts in biomedical applications, a deeper understanding of the molecular level properties of these macromolecules is missing. Herein, we report a simple methodology for the construction of three-dimensional structures of poly-l-lysine dendrigrafts and the subsequent investigation of their structural features using microsecond molecular dynamics simulations. This methodology relies on the encoding of the polymers' experimental characterizations (i.e., composition, degrees of polymerization, branching ratios, charges) into alphanumeric strings that are readable by the Amber simulation package. Such an original approach opens avenues toward the in silico exploration of dendrigrafts and hyperbranched polymers.

Assessing protein-ligand binding modes with computational tools: the case of PDE4B.

In a first step in the discovery of novel potent inhibitor structures for the PDE4B family with limited side effects, we present a protocol to rank newly designed molecules through the estimation of their IC[Formula: see text] values. Our protocol is based on reproducing the linear relationship between the logarithm of experimental IC[Formula: see text] values [[Formula: see text](IC[Formula: see text])] and their calculated binding free energies ([Formula: see text]). From 13 known PDE4B inhibitors, we show here that (1) binding free energies obtained after a docking process by AutoDock are not accurate enough to reproduce this linear relationship; (2) MM-GB/SA post-processing of molecular dynamics (MD) trajectories of the top ranked AutoDock pose improves the linear relationship; (3) by taking into account all representative structures obtained by AutoDock and by averaging MM-GB/SA computations on a series of 40 independent MD trajectories, a linear relationship between [Formula: see text](IC[Formula: see text]) and the lowest [Formula: see text] is achieved with [Formula: see text].

Molecular dynamics simulations of apo, holo, and inactivator bound GABA-at reveal the role of active site residues in PLP dependent enzymes.

The pyridoxal 5-phosphate (PLP) cofactor is a significant organic molecule in medicinal chemistry. It is often found covalently bound to lysine residues in proteins to form PLP dependent enzymes. An example of this family of PLP dependent enzymes is γ-aminobutyric acid aminotransferase (GABA-AT) which is responsible for the degradation of the neurotransmitter GABA. Its inhibition or inactivation can be used to prevent the reduction of GABA concentration in brain which is the source of several neurological disorders. As a test case for PLP dependent enzymes, we have performed molecular dynamics simulations of GABA-AT to reveal the roles of the protein residues and its cofactor. Three different states have been considered: the apoenzyme, the holoenzyme, and the inactive state obtained after the suicide inhibition by vigabatrin. Different protonation states have also been considered for PLP and two key active site residues: Asp298 and His190. Together, 24 independent molecular dynamics trajectories have been simulated for a cumulative total of 2.88 µs. Our results indicate that, unlike in aqueous solution, the PLP pyridine moiety is protonated in GABA-AT. This is a consequence of a pKa shift triggered by a strong charge-charge interaction with an ionic "diad" formed by Asp298 and His190 that would help the activation of the first half-reaction of the catalytic mechanism in GABA-AT: the conversion of PLP to free pyridoxamine phosphate (PMP). In addition, our MD simulations exhibit additional strong hydrogen bond networks between the protein and PLP: the phosphate group is held in place by the donation of at least three hydrogen bonds while the carbonyl oxygen of the pyridine ring interacts with Gln301; Phe181 forms a π-π stacking interaction with the pyridine ring and works as a gate keeper with the assistance of Val300. All these interactions are hypothesized to help maintain free PMP in place inside the protein active site to facilitate the second half-reaction in GABA-AT: the regeneration of PLP-bound GABA-AT (i.e., the holoenzyme). Proteins 2016; 84:875-891. © 2016 Wiley Periodicals, Inc.

Why does Asn71 deamidate faster than Asn15 in the enzyme triosephosphate isomerase? Answers from microsecond molecular dynamics simulation and QM/MM free energy calculations.

Deamidation is the uncatalyzed process by which asparagine or glutamine can be transformed into aspartic acid or glutamic acid, respectively. In its active homodimeric form, mammalian triosephosphate isomerase (TPI) contains two deamidation sites per monomer. Experimental evidence shows that the primary deamidation site (Asn71-Gly72) deamidates faster than the secondary deamidation site (Asn15-Gly16). To evaluate the factors controlling the rates of these two deamidation sites in TPI, we have performed graphics processing unit-enabled microsecond long molecular dynamics simulations of rabbit TPI. The kinetics of asparagine dipeptide and two deamidation sites in mammalian TPI are also investigated using quantum mechanical/molecular mechanical tools with the umbrella sampling technique. Analysis of the simulations has been performed using independent global and local descriptors that can influence the deamidation rates: desolvation effects, backbone acidity, and side chain conformations. Our findings show that all the descriptors add up to favor the primary deamidation site over the secondary one in mammalian TPI: Asn71 deamidates faster because it is more solvent accessible, the adjacent glycine NH backbone acidity is enhanced, and the Asn side chain has a preferential near attack conformation. The crucial impact of the backbone amide acidity of the adjacent glycine on the deamidation rate is shown by kinetic analysis. Our findings also shed light on the effect of high-order structure on deamidation: the deamidation in a small peptide is favored first because of the higher reactivity of the asparagine residue and then because of the stronger stability of the tetrahedral intermediate.

Hydration Effect on Amide I Infrared Bands in Water: An Interpretation Based on an Interaction Energy Decomposition Scheme.

The sensitivity of some infrared bands to the local environment can be exploited to shed light on the structure and the dynamics of biological systems. In particular, the amide I band, which is specifically related to vibrations within the peptide bonds, can give information on the ternary structure of proteins, and can be used as a probe of energy transfer. In this work, we propose a model to quantitatively interpret the frequency shift on the amide I band of a model peptide induced by the formation of hydrogen bonds in the first solvation shell. This method allows us to analyze to what extent the electrostatic interaction, electronic polarization and charge transfer affect the position of the amide I band. The impact of the anharmoniticy of the pontential energy surface on the hydration induced shift is elucidated as well.

Rationalization of the pKa values of alcohols and thiols using atomic charge descriptors and its application to the prediction of amino acid pKa's.

In a first step toward the development of an efficient and accurate protocol to estimate amino acids' pKa's in proteins, we present in this work how to reproduce the pKa's of alcohol and thiol based residues (namely tyrosine, serine, and cysteine) in aqueous solution from the knowledge of the experimental pKa's of phenols, alcohols, and thiols. Our protocol is based on the linear relationship between computed atomic charges of the anionic form of the molecules (being either phenolates, alkoxides, or thiolates) and their respective experimental pKa values. It is tested with different environment approaches (gas phase or continuum solvent-based approaches), with five distinct atomic charge models (Mulliken, Löwdin, NPA, Merz-Kollman, and CHelpG), and with nine different DFT functionals combined with 16 different basis sets. Moreover, the capability of semiempirical methods (AM1, RM1, PM3, and PM6) to also predict pKa's of thiols, phenols, and alcohols is analyzed. From our benchmarks, the best combination to reproduce experimental pKa's is to compute NPA atomic charge using the CPCM model at the B3LYP/3-21G and M062X/6-311G levels for alcohols (R(2) = 0.995) and thiols (R(2) = 0.986), respectively. The applicability of the suggested protocol is tested with tyrosine and cysteine amino acids, and precise pKa predictions are obtained. The stability of the amino acid pKa's with respect to geometrical changes is also tested by MM-MD and DFT-MD calculations. Considering its strong accuracy and its high computational efficiency, these pKa prediction calculations using atomic charges indicate a promising method for predicting amino acids' pKa in a protein environment.

Water interactions with hydrophobic groups: assessment and recalibration of semiempirical molecular orbital methods.

In this work, we present a study of the ability of different semiempirical methods to describe intermolecular interactions in water solution. In particular, we focus on methods based on the Neglect of Diatomic Differential Overlap approximation. Significant improvements of these methods have been reported in the literature in the past years regarding the description of non-covalent interactions. In particular, a broad range of methodologies has been developed to deal with the properties of hydrogen-bonded systems, with varying degrees of success. In contrast, the interactions between water and a molecule containing hydrophobic groups have been little analyzed. Indeed, by considering the potential energy surfaces obtained using different semiempirical Hamiltonians for the intermolecular interactions of model systems, we found that none of the available methods provides an entirely satisfactory description of both hydrophobic and hydrophilic interactions in water. In addition, a vibrational analysis carried out in a model system for these interactions, a methane clathrate cluster, showed that some recent methods cannot be used to carry out studies of vibrational properties. Following a procedure established in our group [M. I. Bernal-Uruchurtu, M. T. C. Martins-Costa, C. Millot, and M. F. Ruiz-López, J. Comput. Chem. 21, 572 (2000); W. Harb, M. I. Bernal-Uruchurtu, and M. F. Ruiz-López, Theor. Chem. Acc. 112, 204 (2004)], we developed new parameters for the core-core interaction terms based on fitting potential energy curves obtained at the MP2 level for our model system. We investigated the transferability of the new parameters to describe a system, having both hydrophilic and hydrophobic groups, interacting with water. We found that only by introducing two different sets of parameters for hydrophilic and hydrophobic hydrogen atom types we are able to match the features of the ab initio calculated properties. Once this assumption is made, a good agreement with the MP2 reference is achieved. The results reported in this work provide therefore a direction for future developments of semiempirical approaches that are still required to investigate chemical processes in biomolecules and in large disordered systems.

Vibrational energy relaxation of the amide I mode of N-methylacetamide in D2O studied through Born-Oppenheimer molecular dynamics.

The vibrational relaxation of the amide I mode of deuterated N-methylacetamide in D2O solution is studied through nonequilibrium simulations using the semiempirical Born-Oppenheimer molecular dynamics (SEBOMD) approach to describe the whole solute-solvent system. Relaxation pathways and lifetimes are determined using the instantaneous normal mode (INM) analysis. The relaxation of the amide I mode is characterized by three different time scales; most of the excess energy (80%) is redistributed through intramolecular vibrational energy redistribution processes, with a smaller contribution (20%) of intermolecular energy flowing into the solvent. The amide II mode is found to contribute modestly (7%) to the relaxation mechanism. The amide I mode and the total vibrational energy decay curves obtained using SEBOMD and INM are in satisfactory agreement with the experimental measurements.

Initiation of the reaction of deamidation in triosephosphate isomerase: investigations by means of molecular dynamics simulations.

Deamidation of asparagine is the spontaneous degradation of this residue into aspartic acid. The kinetics of this slow reaction is mainly dependent on the nature of the adjacent amino acid that follows asparagine in a peptide or protein primary sequence. In the homodimer triosephosphate isomerase (TPI), there are two main deamidation sites per subunit: Asn15-Gly16 and Asn71-Gly72 for which deamidation dynamics are known to be interrelated. In this study, we investigate the initiation of the deamidation reaction in TPI by means of molecular dynamics. Simulations based on classical AMBER force field are performed in a 60 to 90 ns time scale for six distinct samples. Conformational changes, desolvation effects, and hydrogen bond networks are analyzed to interpret the experimental findings and previous quantum mechanical (QM) results. Results that are based on desolvation analysis clarify the assignments in the literature about the different behaviors of two deamidating sites in TPI. Conformational analysis supports findings suggested by QM studies: the most favorable reaction mechanism is the one that yields to succinimide intermediate via one or two step routes. The mechanism leading to the succinimide intermediate most likely involves the formation of a tetrahedral intermediate that is formed either directly from asparagine or via a side chain tautomer intermediate. In all cases, surrounding water molecules are present to assist the reaction.

Peptide binding to ß-cyclodextrins: structure, dynamics, energetics, and electronic effects.

Peptide-cyclodextrin and protein-cyclodextrin host-guest complexes are becoming more and more important for industrial applications, in particular in the fields of pharmaceutical and food chemistry. They have already deserved many experimental investigations although the effect of complex formation in terms of peptide (or protein) structure is not well-known yet. Theoretical calculations represent a unique tool to analyze such effects, and with this aim we have carried out in the present investigation molecular dynamics simulations and combined quantum mechanics-molecular mechanics calculations. We have studied complexes formed between the model Ace-Phe-Nme peptide and the ß-cyclodextrin (ß-CD) macromolecule, and our analysis focuses on the following points: (1) how is the peptide structure modified in going from bulk water to CD environment (backbone torsion angles), (2) which are the main peptide-CD interactions, in particular in terms of hydrogen bonds, (3) which relative peptide-CD orientation is preferred and which are the structural and energetic differences between them, and (4) how the electronic properties of the peptide changes under complex formation. Overall, our calculations show that in the most stable configuration, the backbone chain lies in the narrow rim of the CD. Strong hydrogen bonds form between the H atoms of the peptidic NH groups and oxygen atoms of the secondary OH groups in the CD. These and other (weaker) hydrogen bonds formed by the carbonyl groups reduce considerably the flexibility of the peptide structure, compared to bulk water, and produce a marked increase of the local dipole moment by favoring configurations in which the two C═O bonds point toward the same direction. This effect might have important consequences in terms of the peptide secondary structure, although this hypothesis needs to be tested using larger peptide models.