METHYLCYTISINE ALCALOID POTENTIALLY ACTIVE AGAINST DENGUE VIRUS: A MOLECULAR DOCKING STUDY AND ELECTRONIC STRUCTURAL CHARACTERIZATION

Dengue fever is a serious disease acquired from the Aedes aegypti mosquito present in tropical and subtropical regions, deeply impacting the population's quality of life. Its control requires combating the virus, and the use of substances that do not cause damage to the environment is of fundamental importance. The present work was carried out in silico to perform the structuralelectronic characterization of the alkaloid Methylcytisine, a tricyclic quinolizidine alkaloid that has insecticidal activities, identifying the molecular boundary orbitals and descriptors of global chemical reactivity and assessing the inhibitory potential of methylcytisine on NS5 methyltransferase enzyme dengue virus, as well as identifying possible biological targets in humans. Methylcytisine was geometrically optimized through semi-empirical quantum calculations with thermodynamically more stable conformation, characterizing its structure (atoms, angles and bonds) and its reactivity descriptors. The analysis of the molecular docking simulations showed that methylcytisine is coupled in the same active site of the NS5 enzyme methyltransferase DENV, very similar to the complexed ligand S-adenosyl-L-homocysteine. The intermolecular interactions found for the complex formed and the distance values of the enzyme residues, indicate that methylcytisine has potential application as a new inhibitor of the dengue virus, however it has a high possibility of interaction with human neuronal acetylcholine receptors.


Introduction
In the last 40 years, approximately 3 million cases of dengue shock syndrome and hemorrhagic fevers have been estimated, of which approximately 60 thousand deaths have been recorded [1]. Dengue is a viral disease belonging to the family Flaviviridae, genus Flavivirus, consisting of more than 70 viruses (LEYSSEN et al., 2000), with humans being the main hosts of the dengue virus, a serious febrile disease acquired from the mosquito Aedes aegypti present in tropical and subtropical regions, where the main symptoms are fevers above 38.5 ° C, muscle pain, eye pain, loss of appetite, headache, vomiting and red spots on the body [2].
Dengue can be cured naturally after 15 days, but the hemorrhagic risk impacts on the cardiovascular system, hindering the blood pumping necessary for the body and the functioning of other systems, making the disease extremely serious [3]. Not all drugs can be taken for treatment, as they can aggravate the appearance of bleeding and hemorrhages with the need for new drugs [4]. The Flavivirus genome contains a single open reading frame, which encodes a single polyprotein and which is subsequently cleaved to generate three structural proteins that make up the viral particle: C (capsid), E (envelope) and prM (precursor to the membrane M) and seven nonstructural proteins: NS1, NS2A, NS2B, NS3, NS4A, NS4B and NS5 [5] which are currently considered targets for inhibition of viral replication [6].
The search for substances that have insecticidal activities that do not affect the environment, fits into the concepts of green chemistry, which can be understood as the creation, development and application of chemical products and processes in order to reduce or eliminate the production of products harmful to human health and the environment [7]. In this environment, plant extracts are sources of bioactive substances that can have an effect on insect behavior [8]. Thus, methylcytisine, a tricyclic quinolizidine alkaloid extracted from the roots of plants of the genus Thermopsis, which has several biological activities (hypoglycemic, analgesic and antiinflammatory [9]), and nematicidal activity [10], can be a promising insecticidal agent that fits the concepts of green chemistry.
Currently, for the study of new compounds and their possible biological activities, it is essential to know the structure of the molecule in question, defining the reaction sites, characterizing atomic properties and bonds to then enable the synthesis of the molecule [11] [12]. In this context, molecular modeling, a set of manipulative tools and analysis of complex molecular systems, using in silico methods and theoretical calculations [13] [14], allows the characterization of molecular structures and their properties [15], as well as predicting the mechanism of interaction between molecules and possible biological targets [16] [17].
The present work had as a way to perform the structural-electronic characterization of the alkaloid Methylcytisine in silico by identifying the molecular boundary orbitals and descriptors of global chemical reactivity and to evaluate the inhibitory potential of methylcytisine on NS5 enzyme dengue virus and identify possible biological targets in humans .

Computational Details
The optimization calculations were performed using a computer with 2x Xeon 2.66 Ghz processor -Quad Core, 8 GB Ram DDR2 667ghz ECC, with MAC OSX operating system. Molecular coupling simulations were performed on a personal computer with an Intel® Core TM i7 7700HQ processor, 16 GB RAM, 4G Nvidia® GeForce GTX 1050 Radeon video card and Microsoft Windows 10® operating system. All softwares codes used are free license for academic use.

Obtaining the Structures
To prepare this work, the ChemSpider online repository [http://www.chemspider.com/] was first used to obtain the two-dimensional coordinates of Methylcytisine (ChemSpider ID 204591) and some physico-chemical properties [18] [19].

Structural Optimization
Theoretical calculations, starting from the theory of molecular orbitals, have two premises, the Ab initio method that obtain the results of the modeling through the resolution of all mathematical equations, and the semi-empirical method, where some equations are simplified using experimental data. The semi-empirical method require less computational resources than the Ab initio method, in addition to presenting excellent correlation with experimental results [20]. In this context the Methylcytisine molecule was subjected to structural optimization calculations at the quantum level of semi-empirical, using the Parametric Method 3 (PM3) [21], [22] algorithm, configured so that the Hartree-Fock SCF performs 200 interactions with 10-10 kcal / mol convergence of closed shell. The geometric search was performed with the 10-1 conversion gradient configured to perform a cycle of 1000 interactions with the Quasi-Newton BFGS algorithm line search [13], [23]. All optimization calculations were performed using the ArgusLab® software [24] [25] [26], following the protocol established by Oliveira and collaborators [27] [28].

• Structure Characterization, Frontier Orbitails, MESP and Global Reactivity Molecular Descriptors
The data generated from the geometric optimization were used for the determination of atomic properties, connection, angles, Mulliken population analysis and the electrostatic potential surface map (MESP). Using the Koopmans theorem, the energetic values of the boundary orbitals were used to determine the global reactivity descriptors (   Chemical softness (S) S=1/2 Electrophilicity index (Ω) ω = µ 2 / 2 Chemical potential (µ) µ = -(I+A)/2

Enzyme Preparation
The three-dimensional structure of the dengue virus NS5 methyltransferase was obtained from the Protein Data Bank database (https://www.rcsb.org/structure/1l9k), classified as protein transferase, organism dengue virus type II, expression Escherichia coli, crystallized with the ligands S-adenosyl-L-homocysteine (SAH) and sulfate ions, generated from X-ray diffraction with 2.4 Å resolution and deposited with PDB ID: 1L9K [31]. All water molecules were removed from the methyltransferase structure, reducing interference with the attachment of methylcytisine to the target enzyme. The polar hydrogen atoms missing from the structure were added using AutoDock Tools.

Virtual Screening for Target Classes
In order to assess the possible impacts on the human organism, the methylcistesine molecule was subjected to a virtual screening to identify possible proteins that it can bind to. The algorithm used was that of reverse restraining, which used structural similarities to predict the interaction of possible biological receptors as descriptor, the identifier Homo sapiens was used. For the similarity adjustment, the Tanimoto coefficient was used, vectorized by the Manhattan distance, using the three-dimensional descriptors iej, where d ij is the shortest Manhattan distance between the calculated distances. 20 × 20 in all possible conformations of each molecule [35] calculated using equations 1 and 2.All calculations were performed on the SwissTargetPrediction platform [36].

Two-Dimensional Structure and Physico-Chemical Properties
Technological advances, especially with the increasing accessibility to internet access, have been providing new means of information on molecular data. Methylcytisine (FIGURE 1A), also known as Caulophylline, has a name according to IUPAC, (1R, 9S) -11-methyl-7,11-diazatricyclo [7.3.1.02,7] trideca-2,4-dien-6-one, it is a heterocyclic alkaloid analog from Cytisine (figure 1 B). Regarding the physical-chemical properties, it can be highlight that Methylcytisine in its structure has 3 hydrogen bond acceptor atoms and no donor, has a partition coefficient (LogP) in the order of 0.46, indicating low lipophilicity with strong indications of Low bioaccumulation, this being an important parameter for the study of organic molecules with possible biological actions because it allows to deduce properties such as absorption capacity, half-life, membrane permeation capacity, elimination facilities and permeability capacity ( Table 2).

Geometric Optimization and Structural Properties
Molecular geometry is understood as the organization between atoms due to electronic clouds, which is an important parameter for predicting the polarity of a molecule [37] [38]. When a molecule is removed from a virtual repository or is designed, it does not present itself in its theoretically more stable conformation. In this case, molecular modeling emerges, which is understood as a set of computational tools to build, model, mimic the behavior of molecules [39] [28]. One of the processes used in modeling is the process of geometric optimization which, using theoretical calculations, seeks to hold each atom in its regions with lower levels of potential energy making the molecule thermodynamically more stable [40] [41]. The alkaloid Methylcytisine optimized geometrically ( Figure 02) through semi-empirical quantum calculations, obtaining the structure theoretically close to its native form and potential energy -2286.82885 eV, it is possible to characterize structurally providing atomic properties, of connections and angles. When analyzing the atomic properties (Table 3) it can be notice the presence of variations in the partial charges of atoms of the same nature since the carbon atoms varied from -0.055 to 0.030, the nitrogen atoms from -0.312 to -0.304 and the hydrogens varying from 0.028 to 0.063. It was also possible to characterize all connections (Table 4)

Frontier Orbitals, Global Chemical Reactivity Descriptors
Frontier orbitals are considered fundamental in the study of the chemical reactivity of a molecule and act as intermediates for the stability of structures. The highest energy orbital occupied as HOMO is related to the ionization potential by determining the electron-donor character of the molecule and showing the region where it has the highest electron density. The least energy unoccupied orbital called LUMO is related to the electronic affinity indicating the electronreceptor character showing the lowest electronic density of the molecule [42]. Thus, the higher the HOMO energy, the greater the electron-donor capacity of the molecule and the lower the LUMO energy, the greater the electron acceptance. The variation between the values of the boundary orbitals, called GAP, indicates the energy required for an electron to make the transition. The greater the value of the difference, the lower the molecular reactivity and the greater the molecular stability [43].Methylcytisine presented the energy value -8.47275 eV for the HOMO orbital ( Figure 4A) and having the atoms O1, C2, C3, C4, C5 and N3 as contributors, where all showed symmetry between the positive and negative phases. The LUMO orbital ( Figure 4B) obtained an energy value of -0.12470 eV with atoms O1, C2, C3, C4, C5, C6 and N3 showing symmetry between the positive and negative phases. Starting from the Koopmans theorem, the global reactivity descriptors (generated from the HOMO-LUMO relationship) describe chemical properties that indicate the types of interactions between ligand molecules with biological receptors, since they act as mediators between reactivity and stability. Among these descriptors are: vertical ionization potential (I), electronic affinity (A), electronegativity (χ), chemical hardness (η), chemical softness (S), electronic chemical potential (μ) and electrophilicity index (ω) [29] [44].The ionic potential (I), related to the energy of the HOMO orbital, indicates the strength of an electron's strength when it is attached to an atom. Electronic affinity (A), related to the energy of LUMO, represents the amount of energy released when an atom or molecule receives an electron [45]. Electronegativity is represented by the ability of a molecule to attract electrons from another molecular when they interact, interfering in the dipolar moment and changing molecular properties such as the acidity and basicity of different molecules [27]. Starting from Methylcytisine HOMO-LUMO boundary molecular orbitals, it was possible to determine the global chemical reactivity descriptors (Table 5), where the molecule had a lower GAP value than Cytisine reference, due to the fact that it has lower values of electro affinity and ionic potential than Cytisine. Methylcytisine also showed lower electronegativity and chemical hardness values than Cytsine.

Mulliken Population Analysis
Mulliken atomic charges (Mulliken population analysis) and electrostatic potential surface map Mulliken population analysis divide the charge densities between atoms evenly disregarding electronegativity. When studying the Mulliken charges of Methylcytisine, the existence of varying charges of atoms of the same nature is notorious, where carbon atoms vary from -0.3405 to 0.3035, nitrogen at -0.0954 to 0.0983 and atoms of hydrogens 0.0859 to 0.2280 [46].
The visualization and analysis of charges is possible through the surface map of electrostatic potential, as it allows to characterize electrophilic and nucleophilic regions, showing how complex molecules interact [47] [47]. Through the MESP of Methylcytisine ( Figure 5) it was possible to notice defined regions, since the regions in red mean regions with a high concentration of charges and the regions in white have a low concentration of charge. The regions highlighted in red are due to the presence of a nitrogen atom (N13) and oxygen since they are more electronegative atoms than carbons and hydrogens.

Molecular Docking
Molecular docking positions the ligand in different orientations at the target receptor's active site in order to obtain the best interaction, in which this procedure allows the establishment of a classification between the compounds with the highest and lowest affinity to a given receptor [48] [49] [50] [51]. In order to obtain the best interaction, the molecular coupling positions the ligand in different orientations in the target enzyme, allowing the establishment of a classification between the compounds of greater and lesser affinity to a given receptor [52] [53] [54]. Thus, the molecular coupling simulations between NS5 methyltransferase DENV and the ligand methylcytisine generated energy values of affinity and Root Mean Square Deviation (RMSD) of -6.1 kcal / mol and 1,850 Å respectively, values within the standard of up to 2.0 Å for RMSD [34] and less than -6.0 kcal / mol for affinity [35]. The parameters calculated in the molecular socket of methylcytisine show the affinity of the molecule with the same region of the S-adenosyl-Lhomocysteine ligand complexed in the NS5 enzyme methyltransferase DENV ( Figure 6). The analysis of the distances of methylcytisine with the amino acid residues of the enzyme, highlights the affinity of the ligand with the binding site of S-adenosyl-L-homocysteine (SAH), with methylcytisine being closer to the residues Thr104 (3.8 Å), Lys105 (2.5 Å), Gly106 (2.7 Å) smaller distance compared to SAH, Glu111 and Ile147 (3.7 Å) ( Table 6). Through the analysis of intermolecular interactions of the formed complex, three intermolecular interactions with methylcytisine and residues of NS5 methyltransferase DENV were found, being interactions of the type Carbon Hydrogen Bond with Glu111, Alkyl and Pi-Sigma with Ile147 ( Figure 7).

Virtual Screening for Target Classes
The study of the possible interactions between a substance and the proteins present in the human body is a fundamental factor to determine the viability of using this compound in the environment. It is an initial step for future in vitro and in vivo tests to determine the viability of using the compound. with minimal risks to human health. As for possible classes of biological targets, Methylcistesine was more likely to interact with Ligand-gated ion channel (46.7%), Family AG protein-coupled receptor (26.7%), Protein Kinase (20%) and Enzyme (6.7%) (Figure 8).

Conclusions
Methylcytisine was geometrically optimized through semi-empirical quantum calculations with thermodynamically more stable conformation, characterizing its structure (atoms, angles and bonds) and its reactivity descriptors. The analysis of molecular docking simulations showed that methylcytisine is coupled to the same active site as the NS5 enzyme methyltransferase DENV, in a very similar way to the complexed ligand S-adenosyl-L-homocysteine. The intermolecular interactions found for the complex formed and the distance values of the enzyme residues, indicate that methylcytisine has potential application as a new inhibitor of the dengue virus, however it has a high possibility of interaction with human neuronal acetylcholine receptors.